{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9994227756246392,
  "eval_steps": 500,
  "global_step": 1515,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 2.0000000000000003e-06,
      "loss": 11.3293,
      "step": 10
    },
    {
      "epoch": 0.01,
      "learning_rate": 3.8000000000000005e-06,
      "loss": 10.8095,
      "step": 20
    },
    {
      "epoch": 0.02,
      "learning_rate": 5.8e-06,
      "loss": 11.8008,
      "step": 30
    },
    {
      "epoch": 0.03,
      "learning_rate": 7.800000000000002e-06,
      "loss": 11.1175,
      "step": 40
    },
    {
      "epoch": 0.03,
      "learning_rate": 9.800000000000001e-06,
      "loss": 10.5987,
      "step": 50
    },
    {
      "epoch": 0.04,
      "learning_rate": 1.18e-05,
      "loss": 9.6252,
      "step": 60
    },
    {
      "epoch": 0.05,
      "learning_rate": 1.38e-05,
      "loss": 7.5382,
      "step": 70
    },
    {
      "epoch": 0.05,
      "learning_rate": 1.5600000000000003e-05,
      "loss": 4.7654,
      "step": 80
    },
    {
      "epoch": 0.06,
      "learning_rate": 1.76e-05,
      "loss": 1.7804,
      "step": 90
    },
    {
      "epoch": 0.07,
      "learning_rate": 1.9600000000000002e-05,
      "loss": 0.8864,
      "step": 100
    },
    {
      "epoch": 0.07,
      "learning_rate": 1.999842265960161e-05,
      "loss": 0.6766,
      "step": 110
    },
    {
      "epoch": 0.08,
      "learning_rate": 1.999201556704487e-05,
      "loss": 0.6893,
      "step": 120
    },
    {
      "epoch": 0.09,
      "learning_rate": 1.998068329425523e-05,
      "loss": 0.6321,
      "step": 130
    },
    {
      "epoch": 0.09,
      "learning_rate": 1.996443142704125e-05,
      "loss": 0.6007,
      "step": 140
    },
    {
      "epoch": 0.1,
      "learning_rate": 1.994326797613655e-05,
      "loss": 0.6185,
      "step": 150
    },
    {
      "epoch": 0.11,
      "learning_rate": 1.9917203373251266e-05,
      "loss": 0.6795,
      "step": 160
    },
    {
      "epoch": 0.11,
      "learning_rate": 1.988625046593013e-05,
      "loss": 0.5679,
      "step": 170
    },
    {
      "epoch": 0.12,
      "learning_rate": 1.985042451121975e-05,
      "loss": 0.5809,
      "step": 180
    },
    {
      "epoch": 0.13,
      "learning_rate": 1.9809743168148263e-05,
      "loss": 0.5301,
      "step": 190
    },
    {
      "epoch": 0.13,
      "learning_rate": 1.976422648902097e-05,
      "loss": 0.5804,
      "step": 200
    },
    {
      "epoch": 0.14,
      "learning_rate": 1.971389690953633e-05,
      "loss": 0.5776,
      "step": 210
    },
    {
      "epoch": 0.15,
      "learning_rate": 1.9658779237727157e-05,
      "loss": 0.5567,
      "step": 220
    },
    {
      "epoch": 0.15,
      "learning_rate": 1.9598900641732432e-05,
      "loss": 0.5649,
      "step": 230
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.953429063640583e-05,
      "loss": 0.51,
      "step": 240
    },
    {
      "epoch": 0.16,
      "learning_rate": 1.9464981068767504e-05,
      "loss": 0.5716,
      "step": 250
    },
    {
      "epoch": 0.17,
      "learning_rate": 1.9391006102306304e-05,
      "loss": 0.5424,
      "step": 260
    },
    {
      "epoch": 0.18,
      "learning_rate": 1.931240220014022e-05,
      "loss": 0.4746,
      "step": 270
    },
    {
      "epoch": 0.18,
      "learning_rate": 1.9229208107043274e-05,
      "loss": 0.5063,
      "step": 280
    },
    {
      "epoch": 0.19,
      "learning_rate": 1.914146483034778e-05,
      "loss": 0.4939,
      "step": 290
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.904921561973134e-05,
      "loss": 0.5087,
      "step": 300
    },
    {
      "epoch": 0.2,
      "learning_rate": 1.8952505945898605e-05,
      "loss": 0.5272,
      "step": 310
    },
    {
      "epoch": 0.21,
      "learning_rate": 1.8851383478168206e-05,
      "loss": 0.5222,
      "step": 320
    },
    {
      "epoch": 0.22,
      "learning_rate": 1.8745898060976015e-05,
      "loss": 0.5848,
      "step": 330
    },
    {
      "epoch": 0.22,
      "learning_rate": 1.8636101689306246e-05,
      "loss": 0.5101,
      "step": 340
    },
    {
      "epoch": 0.23,
      "learning_rate": 1.852204848306251e-05,
      "loss": 0.517,
      "step": 350
    },
    {
      "epoch": 0.24,
      "learning_rate": 1.8403794660391496e-05,
      "loss": 0.5125,
      "step": 360
    },
    {
      "epoch": 0.24,
      "learning_rate": 1.8281398509972397e-05,
      "loss": 0.5467,
      "step": 370
    },
    {
      "epoch": 0.25,
      "learning_rate": 1.8154920362285745e-05,
      "loss": 0.5472,
      "step": 380
    },
    {
      "epoch": 0.26,
      "learning_rate": 1.8024422559875815e-05,
      "loss": 0.474,
      "step": 390
    },
    {
      "epoch": 0.26,
      "learning_rate": 1.7889969426621264e-05,
      "loss": 0.4914,
      "step": 400
    },
    {
      "epoch": 0.27,
      "learning_rate": 1.775162723602914e-05,
      "loss": 0.4715,
      "step": 410
    },
    {
      "epoch": 0.28,
      "learning_rate": 1.7609464178567918e-05,
      "loss": 0.5814,
      "step": 420
    },
    {
      "epoch": 0.28,
      "learning_rate": 1.7463550328055613e-05,
      "loss": 0.551,
      "step": 430
    },
    {
      "epoch": 0.29,
      "learning_rate": 1.731395760711959e-05,
      "loss": 0.468,
      "step": 440
    },
    {
      "epoch": 0.3,
      "learning_rate": 1.7160759751745062e-05,
      "loss": 0.5064,
      "step": 450
    },
    {
      "epoch": 0.3,
      "learning_rate": 1.700403227492978e-05,
      "loss": 0.5246,
      "step": 460
    },
    {
      "epoch": 0.31,
      "learning_rate": 1.6843852429462796e-05,
      "loss": 0.4347,
      "step": 470
    },
    {
      "epoch": 0.32,
      "learning_rate": 1.668029916984565e-05,
      "loss": 0.5666,
      "step": 480
    },
    {
      "epoch": 0.32,
      "learning_rate": 1.6513453113374826e-05,
      "loss": 0.5692,
      "step": 490
    },
    {
      "epoch": 0.33,
      "learning_rate": 1.6343396500404503e-05,
      "loss": 0.542,
      "step": 500
    },
    {
      "epoch": 0.34,
      "learning_rate": 1.6170213153809357e-05,
      "loss": 0.4843,
      "step": 510
    },
    {
      "epoch": 0.34,
      "learning_rate": 1.59939884376673e-05,
      "loss": 0.5252,
      "step": 520
    },
    {
      "epoch": 0.35,
      "learning_rate": 1.5814809215182524e-05,
      "loss": 0.4916,
      "step": 530
    },
    {
      "epoch": 0.36,
      "learning_rate": 1.5632763805869625e-05,
      "loss": 0.4334,
      "step": 540
    },
    {
      "epoch": 0.36,
      "learning_rate": 1.5447941942019897e-05,
      "loss": 0.5297,
      "step": 550
    },
    {
      "epoch": 0.37,
      "learning_rate": 1.526043472447123e-05,
      "loss": 0.4612,
      "step": 560
    },
    {
      "epoch": 0.38,
      "learning_rate": 1.507033457770347e-05,
      "loss": 0.4592,
      "step": 570
    },
    {
      "epoch": 0.38,
      "learning_rate": 1.4877735204281323e-05,
      "loss": 0.4032,
      "step": 580
    },
    {
      "epoch": 0.39,
      "learning_rate": 1.4682731538667253e-05,
      "loss": 0.509,
      "step": 590
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.4485419700427203e-05,
      "loss": 0.4458,
      "step": 600
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.4285896946852154e-05,
      "loss": 0.5003,
      "step": 610
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.4084261625018854e-05,
      "loss": 0.4038,
      "step": 620
    },
    {
      "epoch": 0.42,
      "learning_rate": 1.3880613123313431e-05,
      "loss": 0.4024,
      "step": 630
    },
    {
      "epoch": 0.42,
      "learning_rate": 1.3675051822441693e-05,
      "loss": 0.4675,
      "step": 640
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.3467679045950322e-05,
      "loss": 0.4128,
      "step": 650
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.3258597010283302e-05,
      "loss": 0.494,
      "step": 660
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.3047908774398291e-05,
      "loss": 0.4023,
      "step": 670
    },
    {
      "epoch": 0.45,
      "learning_rate": 1.2835718188967623e-05,
      "loss": 0.3956,
      "step": 680
    },
    {
      "epoch": 0.46,
      "learning_rate": 1.2622129845189134e-05,
      "loss": 0.3916,
      "step": 690
    },
    {
      "epoch": 0.46,
      "learning_rate": 1.2407249023231955e-05,
      "loss": 0.4544,
      "step": 700
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.2191181640342664e-05,
      "loss": 0.4408,
      "step": 710
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.1974034198637484e-05,
      "loss": 0.4952,
      "step": 720
    },
    {
      "epoch": 0.48,
      "learning_rate": 1.1755913732606132e-05,
      "loss": 0.4319,
      "step": 730
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.1536927756353302e-05,
      "loss": 0.4349,
      "step": 740
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.1317184210603687e-05,
      "loss": 0.4225,
      "step": 750
    },
    {
      "epoch": 0.5,
      "learning_rate": 1.109679140949677e-05,
      "loss": 0.4221,
      "step": 760
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.0875857987197523e-05,
      "loss": 0.5324,
      "step": 770
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.0654492844349358e-05,
      "loss": 0.5079,
      "step": 780
    },
    {
      "epoch": 0.52,
      "learning_rate": 1.0432805094395759e-05,
      "loss": 0.4659,
      "step": 790
    },
    {
      "epoch": 0.53,
      "learning_rate": 1.021090400979698e-05,
      "loss": 0.4852,
      "step": 800
    },
    {
      "epoch": 0.53,
      "learning_rate": 9.9888989681684e-06,
      "loss": 0.5068,
      "step": 810
    },
    {
      "epoch": 0.54,
      "learning_rate": 9.766899398367015e-06,
      "loss": 0.4887,
      "step": 820
    },
    {
      "epoch": 0.55,
      "learning_rate": 9.545014726552706e-06,
      "loss": 0.4185,
      "step": 830
    },
    {
      "epoch": 0.55,
      "learning_rate": 9.323354322250787e-06,
      "loss": 0.4555,
      "step": 840
    },
    {
      "epoch": 0.56,
      "learning_rate": 9.102027444442539e-06,
      "loss": 0.4548,
      "step": 850
    },
    {
      "epoch": 0.57,
      "learning_rate": 8.881143187710148e-06,
      "loss": 0.4426,
      "step": 860
    },
    {
      "epoch": 0.57,
      "learning_rate": 8.660810428462762e-06,
      "loss": 0.4543,
      "step": 870
    },
    {
      "epoch": 0.58,
      "learning_rate": 8.441137771270017e-06,
      "loss": 0.3691,
      "step": 880
    },
    {
      "epoch": 0.59,
      "learning_rate": 8.222233495329605e-06,
      "loss": 0.4368,
      "step": 890
    },
    {
      "epoch": 0.59,
      "learning_rate": 8.004205501095199e-06,
      "loss": 0.4426,
      "step": 900
    },
    {
      "epoch": 0.6,
      "learning_rate": 7.78716125709106e-06,
      "loss": 0.4359,
      "step": 910
    },
    {
      "epoch": 0.61,
      "learning_rate": 7.571207746939571e-06,
      "loss": 0.443,
      "step": 920
    },
    {
      "epoch": 0.61,
      "learning_rate": 7.35645141662775e-06,
      "loss": 0.4758,
      "step": 930
    },
    {
      "epoch": 0.62,
      "learning_rate": 7.142998122038824e-06,
      "loss": 0.4666,
      "step": 940
    },
    {
      "epoch": 0.63,
      "learning_rate": 6.930953076774594e-06,
      "loss": 0.4006,
      "step": 950
    },
    {
      "epoch": 0.63,
      "learning_rate": 6.720420800294467e-06,
      "loss": 0.4319,
      "step": 960
    },
    {
      "epoch": 0.64,
      "learning_rate": 6.511505066396585e-06,
      "loss": 0.4168,
      "step": 970
    },
    {
      "epoch": 0.65,
      "learning_rate": 6.3043088520665305e-06,
      "loss": 0.4459,
      "step": 980
    },
    {
      "epoch": 0.65,
      "learning_rate": 6.098934286718755e-06,
      "loss": 0.4616,
      "step": 990
    },
    {
      "epoch": 0.66,
      "learning_rate": 5.89548260185584e-06,
      "loss": 0.4184,
      "step": 1000
    },
    {
      "epoch": 0.67,
      "learning_rate": 5.694054081170292e-06,
      "loss": 0.4148,
      "step": 1010
    },
    {
      "epoch": 0.67,
      "learning_rate": 5.494748011113545e-06,
      "loss": 0.4819,
      "step": 1020
    },
    {
      "epoch": 0.68,
      "learning_rate": 5.297662631956552e-06,
      "loss": 0.4848,
      "step": 1030
    },
    {
      "epoch": 0.69,
      "learning_rate": 5.102895089365969e-06,
      "loss": 0.3852,
      "step": 1040
    },
    {
      "epoch": 0.69,
      "learning_rate": 4.910541386519974e-06,
      "loss": 0.5062,
      "step": 1050
    },
    {
      "epoch": 0.7,
      "learning_rate": 4.720696336787162e-06,
      "loss": 0.4424,
      "step": 1060
    },
    {
      "epoch": 0.71,
      "learning_rate": 4.533453516991947e-06,
      "loss": 0.4368,
      "step": 1070
    },
    {
      "epoch": 0.71,
      "learning_rate": 4.348905221289454e-06,
      "loss": 0.473,
      "step": 1080
    },
    {
      "epoch": 0.72,
      "learning_rate": 4.167142415672681e-06,
      "loss": 0.4313,
      "step": 1090
    },
    {
      "epoch": 0.73,
      "learning_rate": 3.988254693134277e-06,
      "loss": 0.4915,
      "step": 1100
    },
    {
      "epoch": 0.73,
      "learning_rate": 3.8123302295051334e-06,
      "loss": 0.4222,
      "step": 1110
    },
    {
      "epoch": 0.74,
      "learning_rate": 3.639455739991512e-06,
      "loss": 0.4552,
      "step": 1120
    },
    {
      "epoch": 0.75,
      "learning_rate": 3.4697164364320735e-06,
      "loss": 0.564,
      "step": 1130
    },
    {
      "epoch": 0.75,
      "learning_rate": 3.3031959852960216e-06,
      "loss": 0.5002,
      "step": 1140
    },
    {
      "epoch": 0.76,
      "learning_rate": 3.1399764664428987e-06,
      "loss": 0.4253,
      "step": 1150
    },
    {
      "epoch": 0.77,
      "learning_rate": 2.9801383326644827e-06,
      "loss": 0.4564,
      "step": 1160
    },
    {
      "epoch": 0.77,
      "learning_rate": 2.8237603700286754e-06,
      "loss": 0.504,
      "step": 1170
    },
    {
      "epoch": 0.78,
      "learning_rate": 2.6709196590449406e-06,
      "loss": 0.5243,
      "step": 1180
    },
    {
      "epoch": 0.79,
      "learning_rate": 2.5216915366704153e-06,
      "loss": 0.4597,
      "step": 1190
    },
    {
      "epoch": 0.79,
      "learning_rate": 2.3761495591754602e-06,
      "loss": 0.4028,
      "step": 1200
    },
    {
      "epoch": 0.8,
      "learning_rate": 2.2343654658869296e-06,
      "loss": 0.4516,
      "step": 1210
    },
    {
      "epoch": 0.8,
      "learning_rate": 2.0964091438269953e-06,
      "loss": 0.4573,
      "step": 1220
    },
    {
      "epoch": 0.81,
      "learning_rate": 1.9623485932650665e-06,
      "loss": 0.4844,
      "step": 1230
    },
    {
      "epoch": 0.82,
      "learning_rate": 1.832249894199637e-06,
      "loss": 0.5323,
      "step": 1240
    },
    {
      "epoch": 0.82,
      "learning_rate": 1.7061771737867007e-06,
      "loss": 0.4656,
      "step": 1250
    },
    {
      "epoch": 0.83,
      "learning_rate": 1.5841925747307285e-06,
      "loss": 0.507,
      "step": 1260
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.4663562246538231e-06,
      "loss": 0.513,
      "step": 1270
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.3527262064580925e-06,
      "loss": 0.3735,
      "step": 1280
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.2433585296959405e-06,
      "loss": 0.5576,
      "step": 1290
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.1383071029622926e-06,
      "loss": 0.4359,
      "step": 1300
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.0376237073224304e-06,
      "loss": 0.4944,
      "step": 1310
    },
    {
      "epoch": 0.87,
      "learning_rate": 9.413579707885101e-07,
      "loss": 0.4297,
      "step": 1320
    },
    {
      "epoch": 0.88,
      "learning_rate": 8.49557343857329e-07,
      "loss": 0.5633,
      "step": 1330
    },
    {
      "epoch": 0.88,
      "learning_rate": 7.622670761214323e-07,
      "loss": 0.4062,
      "step": 1340
    },
    {
      "epoch": 0.89,
      "learning_rate": 6.795301939650656e-07,
      "loss": 0.4329,
      "step": 1350
    },
    {
      "epoch": 0.9,
      "learning_rate": 6.013874793559737e-07,
      "loss": 0.479,
      "step": 1360
    },
    {
      "epoch": 0.9,
      "learning_rate": 5.278774497434857e-07,
      "loss": 0.4258,
      "step": 1370
    },
    {
      "epoch": 0.91,
      "learning_rate": 4.590363390728314e-07,
      "loss": 0.5493,
      "step": 1380
    },
    {
      "epoch": 0.92,
      "learning_rate": 3.948980799250002e-07,
      "loss": 0.5105,
      "step": 1390
    },
    {
      "epoch": 0.92,
      "learning_rate": 3.3549428679098274e-07,
      "loss": 0.4404,
      "step": 1400
    },
    {
      "epoch": 0.93,
      "learning_rate": 2.8085424048862165e-07,
      "loss": 0.484,
      "step": 1410
    },
    {
      "epoch": 0.94,
      "learning_rate": 2.3100487372975987e-07,
      "loss": 0.37,
      "step": 1420
    },
    {
      "epoch": 0.94,
      "learning_rate": 1.859707578447878e-07,
      "loss": 0.4036,
      "step": 1430
    },
    {
      "epoch": 0.95,
      "learning_rate": 1.4577409067115466e-07,
      "loss": 0.4715,
      "step": 1440
    },
    {
      "epoch": 0.96,
      "learning_rate": 1.1043468561180326e-07,
      "loss": 0.5,
      "step": 1450
    },
    {
      "epoch": 0.96,
      "learning_rate": 7.99699618688976e-08,
      "loss": 0.4995,
      "step": 1460
    },
    {
      "epoch": 0.97,
      "learning_rate": 5.439493585771449e-08,
      "loss": 0.491,
      "step": 1470
    },
    {
      "epoch": 0.98,
      "learning_rate": 3.372221380486651e-08,
      "loss": 0.5358,
      "step": 1480
    },
    {
      "epoch": 0.98,
      "learning_rate": 1.7961985534552528e-08,
      "loss": 0.4628,
      "step": 1490
    },
    {
      "epoch": 0.99,
      "learning_rate": 7.122019445873207e-09,
      "loss": 0.4376,
      "step": 1500
    },
    {
      "epoch": 1.0,
      "learning_rate": 1.2076586836973214e-09,
      "loss": 0.4577,
      "step": 1510
    },
    {
      "epoch": 1.0,
      "step": 1515,
      "total_flos": 1.1497394350374912e+18,
      "train_loss": 0.9827296845590321,
      "train_runtime": 12477.9588,
      "train_samples_per_second": 3.887,
      "train_steps_per_second": 0.121
    }
  ],
  "logging_steps": 10,
  "max_steps": 1515,
  "num_train_epochs": 1,
  "save_steps": 200000,
  "total_flos": 1.1497394350374912e+18,
  "trial_name": null,
  "trial_params": null
}