{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9995754805461492,
  "eval_steps": 500,
  "global_step": 1582,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.01,
      "learning_rate": 2.999810713827213e-05,
      "loss": 3.8169,
      "step": 10
    },
    {
      "epoch": 0.01,
      "learning_rate": 2.999041820624981e-05,
      "loss": 3.4444,
      "step": 20
    },
    {
      "epoch": 0.02,
      "learning_rate": 2.9976817929807542e-05,
      "loss": 3.3706,
      "step": 30
    },
    {
      "epoch": 0.03,
      "learning_rate": 2.995731167209911e-05,
      "loss": 3.3325,
      "step": 40
    },
    {
      "epoch": 0.03,
      "learning_rate": 2.9931907125251988e-05,
      "loss": 3.2825,
      "step": 50
    },
    {
      "epoch": 0.04,
      "learning_rate": 2.9900614307334e-05,
      "loss": 3.273,
      "step": 60
    },
    {
      "epoch": 0.04,
      "learning_rate": 2.986344555840277e-05,
      "loss": 3.2433,
      "step": 70
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.982041553563955e-05,
      "loss": 3.2409,
      "step": 80
    },
    {
      "epoch": 0.06,
      "learning_rate": 2.977154120756926e-05,
      "loss": 3.2218,
      "step": 90
    },
    {
      "epoch": 0.06,
      "learning_rate": 2.9716841847369106e-05,
      "loss": 3.2128,
      "step": 100
    },
    {
      "epoch": 0.07,
      "learning_rate": 2.9656339025268374e-05,
      "loss": 3.2057,
      "step": 110
    },
    {
      "epoch": 0.08,
      "learning_rate": 2.959005660004237e-05,
      "loss": 3.1869,
      "step": 120
    },
    {
      "epoch": 0.08,
      "learning_rate": 2.9518020709603938e-05,
      "loss": 3.1853,
      "step": 130
    },
    {
      "epoch": 0.09,
      "learning_rate": 2.9440259760696174e-05,
      "loss": 3.1747,
      "step": 140
    },
    {
      "epoch": 0.09,
      "learning_rate": 2.935680441769049e-05,
      "loss": 3.1569,
      "step": 150
    },
    {
      "epoch": 0.1,
      "learning_rate": 2.9267687590494364e-05,
      "loss": 3.147,
      "step": 160
    },
    {
      "epoch": 0.11,
      "learning_rate": 2.9172944421573587e-05,
      "loss": 3.1393,
      "step": 170
    },
    {
      "epoch": 0.11,
      "learning_rate": 2.9072612272094165e-05,
      "loss": 3.1262,
      "step": 180
    },
    {
      "epoch": 0.12,
      "learning_rate": 2.8966730707189218e-05,
      "loss": 3.1286,
      "step": 190
    },
    {
      "epoch": 0.13,
      "learning_rate": 2.885534148035684e-05,
      "loss": 3.1145,
      "step": 200
    },
    {
      "epoch": 0.13,
      "learning_rate": 2.8738488516994925e-05,
      "loss": 3.1037,
      "step": 210
    },
    {
      "epoch": 0.14,
      "learning_rate": 2.8616217897079593e-05,
      "loss": 3.1064,
      "step": 220
    },
    {
      "epoch": 0.15,
      "learning_rate": 2.8488577836993934e-05,
      "loss": 3.1016,
      "step": 230
    },
    {
      "epoch": 0.15,
      "learning_rate": 2.835561867051426e-05,
      "loss": 3.0887,
      "step": 240
    },
    {
      "epoch": 0.16,
      "learning_rate": 2.821739282896143e-05,
      "loss": 3.0969,
      "step": 250
    },
    {
      "epoch": 0.16,
      "learning_rate": 2.8073954820525014e-05,
      "loss": 3.0809,
      "step": 260
    },
    {
      "epoch": 0.17,
      "learning_rate": 2.792536120876842e-05,
      "loss": 3.0821,
      "step": 270
    },
    {
      "epoch": 0.18,
      "learning_rate": 2.7771670590323548e-05,
      "loss": 3.0799,
      "step": 280
    },
    {
      "epoch": 0.18,
      "learning_rate": 2.7612943571783705e-05,
      "loss": 3.0651,
      "step": 290
    },
    {
      "epoch": 0.19,
      "learning_rate": 2.7449242745803914e-05,
      "loss": 3.0548,
      "step": 300
    },
    {
      "epoch": 0.2,
      "learning_rate": 2.7280632666418013e-05,
      "loss": 3.0592,
      "step": 310
    },
    {
      "epoch": 0.2,
      "learning_rate": 2.710717982358233e-05,
      "loss": 3.0647,
      "step": 320
    },
    {
      "epoch": 0.21,
      "learning_rate": 2.6928952616955944e-05,
      "loss": 3.0397,
      "step": 330
    },
    {
      "epoch": 0.21,
      "learning_rate": 2.674602132892783e-05,
      "loss": 3.0382,
      "step": 340
    },
    {
      "epoch": 0.22,
      "learning_rate": 2.655845809690162e-05,
      "loss": 3.0304,
      "step": 350
    },
    {
      "epoch": 0.23,
      "learning_rate": 2.636633688484882e-05,
      "loss": 3.0435,
      "step": 360
    },
    {
      "epoch": 0.23,
      "learning_rate": 2.616973345414173e-05,
      "loss": 3.043,
      "step": 370
    },
    {
      "epoch": 0.24,
      "learning_rate": 2.598902211115875e-05,
      "loss": 3.0378,
      "step": 380
    },
    {
      "epoch": 0.25,
      "learning_rate": 2.5784117493354177e-05,
      "loss": 3.0303,
      "step": 390
    },
    {
      "epoch": 0.25,
      "learning_rate": 2.5574960250179765e-05,
      "loss": 3.0149,
      "step": 400
    },
    {
      "epoch": 0.26,
      "learning_rate": 2.5361632861022366e-05,
      "loss": 3.0235,
      "step": 410
    },
    {
      "epoch": 0.27,
      "learning_rate": 2.5144219449730562e-05,
      "loss": 3.0245,
      "step": 420
    },
    {
      "epoch": 0.27,
      "learning_rate": 2.4922805751441174e-05,
      "loss": 3.0177,
      "step": 430
    },
    {
      "epoch": 0.28,
      "learning_rate": 2.469747907877028e-05,
      "loss": 3.0119,
      "step": 440
    },
    {
      "epoch": 0.28,
      "learning_rate": 2.4468328287382274e-05,
      "loss": 3.0069,
      "step": 450
    },
    {
      "epoch": 0.29,
      "learning_rate": 2.423544374095036e-05,
      "loss": 2.9941,
      "step": 460
    },
    {
      "epoch": 0.3,
      "learning_rate": 2.3998917275522408e-05,
      "loss": 3.0015,
      "step": 470
    },
    {
      "epoch": 0.3,
      "learning_rate": 2.37588421633062e-05,
      "loss": 3.0088,
      "step": 480
    },
    {
      "epoch": 0.31,
      "learning_rate": 2.35153130758883e-05,
      "loss": 3.0014,
      "step": 490
    },
    {
      "epoch": 0.32,
      "learning_rate": 2.3268426046901153e-05,
      "loss": 2.9899,
      "step": 500
    },
    {
      "epoch": 0.32,
      "learning_rate": 2.3018278434152984e-05,
      "loss": 2.9969,
      "step": 510
    },
    {
      "epoch": 0.33,
      "learning_rate": 2.2764968881235625e-05,
      "loss": 3.0184,
      "step": 520
    },
    {
      "epoch": 0.33,
      "learning_rate": 2.2508597278625205e-05,
      "loss": 3.003,
      "step": 530
    },
    {
      "epoch": 0.34,
      "learning_rate": 2.2249264724291235e-05,
      "loss": 2.9918,
      "step": 540
    },
    {
      "epoch": 0.35,
      "learning_rate": 2.1987073483829422e-05,
      "loss": 2.9868,
      "step": 550
    },
    {
      "epoch": 0.35,
      "learning_rate": 2.172212695013415e-05,
      "loss": 2.9815,
      "step": 560
    },
    {
      "epoch": 0.36,
      "learning_rate": 2.1454529602626336e-05,
      "loss": 2.9837,
      "step": 570
    },
    {
      "epoch": 0.37,
      "learning_rate": 2.1184386966052872e-05,
      "loss": 2.9728,
      "step": 580
    },
    {
      "epoch": 0.37,
      "learning_rate": 2.0911805568873825e-05,
      "loss": 2.9696,
      "step": 590
    },
    {
      "epoch": 0.38,
      "learning_rate": 2.063689290125385e-05,
      "loss": 2.9709,
      "step": 600
    },
    {
      "epoch": 0.39,
      "learning_rate": 2.0359757372674344e-05,
      "loss": 2.97,
      "step": 610
    },
    {
      "epoch": 0.39,
      "learning_rate": 2.00805082691831e-05,
      "loss": 2.9621,
      "step": 620
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.9799255710298257e-05,
      "loss": 2.9608,
      "step": 630
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.951611060558363e-05,
      "loss": 2.9673,
      "step": 640
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.9231184610912436e-05,
      "loss": 2.9609,
      "step": 650
    },
    {
      "epoch": 0.42,
      "learning_rate": 1.8944590084436768e-05,
      "loss": 2.9596,
      "step": 660
    },
    {
      "epoch": 0.42,
      "learning_rate": 1.86564400422801e-05,
      "loss": 2.9589,
      "step": 670
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.8366848113970322e-05,
      "loss": 2.9513,
      "step": 680
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.8075928497630936e-05,
      "loss": 2.9559,
      "step": 690
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.7783795914947944e-05,
      "loss": 2.9377,
      "step": 700
    },
    {
      "epoch": 0.45,
      "learning_rate": 1.7490565565930382e-05,
      "loss": 2.9469,
      "step": 710
    },
    {
      "epoch": 0.45,
      "learning_rate": 1.7196353083482102e-05,
      "loss": 2.9454,
      "step": 720
    },
    {
      "epoch": 0.46,
      "learning_rate": 1.6901274487802977e-05,
      "loss": 2.9426,
      "step": 730
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.66054461406373e-05,
      "loss": 2.9504,
      "step": 740
    },
    {
      "epoch": 0.47,
      "learning_rate": 1.6308984699387494e-05,
      "loss": 2.9388,
      "step": 750
    },
    {
      "epoch": 0.48,
      "learning_rate": 1.6012007071111277e-05,
      "loss": 2.9356,
      "step": 760
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.5714630366420347e-05,
      "loss": 2.938,
      "step": 770
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.541697185329881e-05,
      "loss": 2.9352,
      "step": 780
    },
    {
      "epoch": 0.5,
      "learning_rate": 1.5119148910859536e-05,
      "loss": 2.933,
      "step": 790
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.4821278983056699e-05,
      "loss": 2.9314,
      "step": 800
    },
    {
      "epoch": 0.51,
      "learning_rate": 1.4523479532372764e-05,
      "loss": 2.9319,
      "step": 810
    },
    {
      "epoch": 0.52,
      "learning_rate": 1.4225867993498119e-05,
      "loss": 2.9238,
      "step": 820
    },
    {
      "epoch": 0.52,
      "learning_rate": 1.3928561727021752e-05,
      "loss": 2.9338,
      "step": 830
    },
    {
      "epoch": 0.53,
      "learning_rate": 1.363167797315105e-05,
      "loss": 2.9274,
      "step": 840
    },
    {
      "epoch": 0.54,
      "learning_rate": 1.3335333805479126e-05,
      "loss": 2.9247,
      "step": 850
    },
    {
      "epoch": 0.54,
      "learning_rate": 1.3039646084817877e-05,
      "loss": 2.9243,
      "step": 860
    },
    {
      "epoch": 0.55,
      "learning_rate": 1.2744731413114838e-05,
      "loss": 2.9208,
      "step": 870
    },
    {
      "epoch": 0.56,
      "learning_rate": 1.2450706087472275e-05,
      "loss": 2.9209,
      "step": 880
    },
    {
      "epoch": 0.56,
      "learning_rate": 1.215768605428635e-05,
      "loss": 2.9274,
      "step": 890
    },
    {
      "epoch": 0.57,
      "learning_rate": 1.1865786863524683e-05,
      "loss": 2.923,
      "step": 900
    },
    {
      "epoch": 0.57,
      "learning_rate": 1.157512362316014e-05,
      "loss": 2.9202,
      "step": 910
    },
    {
      "epoch": 0.58,
      "learning_rate": 1.1285810953779057e-05,
      "loss": 2.9186,
      "step": 920
    },
    {
      "epoch": 0.59,
      "learning_rate": 1.0997962943381529e-05,
      "loss": 2.9166,
      "step": 930
    },
    {
      "epoch": 0.59,
      "learning_rate": 1.0711693102391794e-05,
      "loss": 2.9149,
      "step": 940
    },
    {
      "epoch": 0.6,
      "learning_rate": 1.0427114318896343e-05,
      "loss": 2.9076,
      "step": 950
    },
    {
      "epoch": 0.61,
      "learning_rate": 1.014433881412749e-05,
      "loss": 2.9179,
      "step": 960
    },
    {
      "epoch": 0.61,
      "learning_rate": 9.863478098209844e-06,
      "loss": 2.9127,
      "step": 970
    },
    {
      "epoch": 0.62,
      "learning_rate": 9.584642926187264e-06,
      "loss": 2.9162,
      "step": 980
    },
    {
      "epoch": 0.63,
      "learning_rate": 9.307943254347521e-06,
      "loss": 2.9087,
      "step": 990
    },
    {
      "epoch": 0.63,
      "learning_rate": 9.033488196861998e-06,
      "loss": 2.9007,
      "step": 1000
    },
    {
      "epoch": 0.64,
      "learning_rate": 8.76138598275741e-06,
      "loss": 2.9102,
      "step": 1010
    },
    {
      "epoch": 0.64,
      "learning_rate": 8.491743913236629e-06,
      "loss": 2.9006,
      "step": 1020
    },
    {
      "epoch": 0.65,
      "learning_rate": 8.224668319365347e-06,
      "loss": 2.9061,
      "step": 1030
    },
    {
      "epoch": 0.66,
      "learning_rate": 7.960264520141317e-06,
      "loss": 2.9052,
      "step": 1040
    },
    {
      "epoch": 0.66,
      "learning_rate": 7.698636780962729e-06,
      "loss": 2.9017,
      "step": 1050
    },
    {
      "epoch": 0.67,
      "learning_rate": 7.439888272512004e-06,
      "loss": 2.8945,
      "step": 1060
    },
    {
      "epoch": 0.68,
      "learning_rate": 7.184121030071315e-06,
      "loss": 2.8982,
      "step": 1070
    },
    {
      "epoch": 0.68,
      "learning_rate": 6.931435913285836e-06,
      "loss": 2.9084,
      "step": 1080
    },
    {
      "epoch": 0.69,
      "learning_rate": 6.681932566390615e-06,
      "loss": 2.8975,
      "step": 1090
    },
    {
      "epoch": 0.7,
      "learning_rate": 6.4357093789167005e-06,
      "loss": 2.9089,
      "step": 1100
    },
    {
      "epoch": 0.7,
      "learning_rate": 6.192863446892048e-06,
      "loss": 2.9079,
      "step": 1110
    },
    {
      "epoch": 0.71,
      "learning_rate": 5.953490534552541e-06,
      "loss": 2.9077,
      "step": 1120
    },
    {
      "epoch": 0.71,
      "learning_rate": 5.71768503657819e-06,
      "loss": 2.8958,
      "step": 1130
    },
    {
      "epoch": 0.72,
      "learning_rate": 5.485539940869361e-06,
      "loss": 2.8987,
      "step": 1140
    },
    {
      "epoch": 0.73,
      "learning_rate": 5.2571467918777955e-06,
      "loss": 2.892,
      "step": 1150
    },
    {
      "epoch": 0.73,
      "learning_rate": 5.032595654506847e-06,
      "loss": 2.8925,
      "step": 1160
    },
    {
      "epoch": 0.74,
      "learning_rate": 4.811975078595121e-06,
      "loss": 2.8975,
      "step": 1170
    },
    {
      "epoch": 0.75,
      "learning_rate": 4.5953720639976185e-06,
      "loss": 2.9005,
      "step": 1180
    },
    {
      "epoch": 0.75,
      "learning_rate": 4.382872026278071e-06,
      "loss": 2.893,
      "step": 1190
    },
    {
      "epoch": 0.76,
      "learning_rate": 4.174558763026048e-06,
      "loss": 2.8926,
      "step": 1200
    },
    {
      "epoch": 0.76,
      "learning_rate": 3.970514420812069e-06,
      "loss": 2.8865,
      "step": 1210
    },
    {
      "epoch": 0.77,
      "learning_rate": 3.770819462793801e-06,
      "loss": 2.8992,
      "step": 1220
    },
    {
      "epoch": 0.78,
      "learning_rate": 3.5755526369861207e-06,
      "loss": 2.8841,
      "step": 1230
    },
    {
      "epoch": 0.78,
      "learning_rate": 3.3847909452074768e-06,
      "loss": 2.8843,
      "step": 1240
    },
    {
      "epoch": 0.79,
      "learning_rate": 3.1986096127148724e-06,
      "loss": 2.8883,
      "step": 1250
    },
    {
      "epoch": 0.8,
      "learning_rate": 3.0170820585394327e-06,
      "loss": 2.8919,
      "step": 1260
    },
    {
      "epoch": 0.8,
      "learning_rate": 2.8402798665342412e-06,
      "loss": 2.8855,
      "step": 1270
    },
    {
      "epoch": 0.81,
      "learning_rate": 2.668272757145839e-06,
      "loss": 2.8964,
      "step": 1280
    },
    {
      "epoch": 0.82,
      "learning_rate": 2.5011285599205554e-06,
      "loss": 2.8939,
      "step": 1290
    },
    {
      "epoch": 0.82,
      "learning_rate": 2.3389131867565084e-06,
      "loss": 2.8852,
      "step": 1300
    },
    {
      "epoch": 0.83,
      "learning_rate": 2.18169060591181e-06,
      "loss": 2.8905,
      "step": 1310
    },
    {
      "epoch": 0.83,
      "learning_rate": 2.0295228167792224e-06,
      "loss": 2.8827,
      "step": 1320
    },
    {
      "epoch": 0.84,
      "learning_rate": 1.8824698254372108e-06,
      "loss": 2.8844,
      "step": 1330
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.7405896209870665e-06,
      "loss": 2.8878,
      "step": 1340
    },
    {
      "epoch": 0.85,
      "learning_rate": 1.6039381526854018e-06,
      "loss": 2.8825,
      "step": 1350
    },
    {
      "epoch": 0.86,
      "learning_rate": 1.4725693078810238e-06,
      "loss": 2.8849,
      "step": 1360
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.3465348907649138e-06,
      "loss": 2.8894,
      "step": 1370
    },
    {
      "epoch": 0.87,
      "learning_rate": 1.2258846019416892e-06,
      "loss": 2.882,
      "step": 1380
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.110666018830594e-06,
      "loss": 2.8784,
      "step": 1390
    },
    {
      "epoch": 0.88,
      "learning_rate": 1.0009245769037412e-06,
      "loss": 2.8918,
      "step": 1400
    },
    {
      "epoch": 0.89,
      "learning_rate": 8.967035517690147e-07,
      "loss": 2.887,
      "step": 1410
    },
    {
      "epoch": 0.9,
      "learning_rate": 7.980440421047119e-07,
      "loss": 2.8869,
      "step": 1420
    },
    {
      "epoch": 0.9,
      "learning_rate": 7.049849534526192e-07,
      "loss": 2.8779,
      "step": 1430
    },
    {
      "epoch": 0.91,
      "learning_rate": 6.175629828759482e-07,
      "loss": 2.885,
      "step": 1440
    },
    {
      "epoch": 0.92,
      "learning_rate": 5.358126044881461e-07,
      "loss": 2.8866,
      "step": 1450
    },
    {
      "epoch": 0.92,
      "learning_rate": 4.5976605585834055e-07,
      "loss": 2.8828,
      "step": 1460
    },
    {
      "epoch": 0.93,
      "learning_rate": 3.894533252987098e-07,
      "loss": 2.8861,
      "step": 1470
    },
    {
      "epoch": 0.94,
      "learning_rate": 3.2490214003885963e-07,
      "loss": 2.8815,
      "step": 1480
    },
    {
      "epoch": 0.94,
      "learning_rate": 2.661379552918142e-07,
      "loss": 2.8752,
      "step": 1490
    },
    {
      "epoch": 0.95,
      "learning_rate": 2.1318394421597553e-07,
      "loss": 2.8789,
      "step": 1500
    },
    {
      "epoch": 0.95,
      "learning_rate": 1.660609887769804e-07,
      "loss": 2.8841,
      "step": 1510
    },
    {
      "epoch": 0.96,
      "learning_rate": 1.2478767151308025e-07,
      "loss": 2.88,
      "step": 1520
    },
    {
      "epoch": 0.97,
      "learning_rate": 8.938026820726641e-08,
      "loss": 2.8744,
      "step": 1530
    },
    {
      "epoch": 0.97,
      "learning_rate": 5.985274146905917e-08,
      "loss": 2.8835,
      "step": 1540
    },
    {
      "epoch": 0.98,
      "learning_rate": 3.621673522847035e-08,
      "loss": 2.8893,
      "step": 1550
    },
    {
      "epoch": 0.99,
      "learning_rate": 1.848157014431473e-08,
      "loss": 2.8894,
      "step": 1560
    },
    {
      "epoch": 0.99,
      "learning_rate": 6.65423992868841e-09,
      "loss": 2.885,
      "step": 1570
    },
    {
      "epoch": 1.0,
      "learning_rate": 7.394085890616298e-10,
      "loss": 2.8879,
      "step": 1580
    },
    {
      "epoch": 1.0,
      "step": 1582,
      "total_flos": 7.0894081521926275e+19,
      "train_loss": 2.982507556789895,
      "train_runtime": 289179.519,
      "train_samples_per_second": 5.604,
      "train_steps_per_second": 0.005
    }
  ],
  "logging_steps": 10,
  "max_steps": 1582,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 1000,
  "total_flos": 7.0894081521926275e+19,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}