{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 25,
  "global_step": 1030,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "learning_rate": 4.999712746672285e-05,
      "loss": 3.1322,
      "step": 5
    },
    {
      "epoch": 0.05,
      "learning_rate": 4.998844107474698e-05,
      "loss": 0.2475,
      "step": 10
    },
    {
      "epoch": 0.07,
      "learning_rate": 4.997394273973373e-05,
      "loss": 0.1435,
      "step": 15
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.9953635833782084e-05,
      "loss": 0.1028,
      "step": 20
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.992752507997904e-05,
      "loss": 0.0952,
      "step": 25
    },
    {
      "epoch": 0.12,
      "eval_loss": 0.09023847430944443,
      "eval_runtime": 0.122,
      "eval_samples_per_second": 3368.392,
      "eval_steps_per_second": 106.543,
      "step": 25
    },
    {
      "epoch": 0.15,
      "learning_rate": 4.989561655130103e-05,
      "loss": 0.1171,
      "step": 30
    },
    {
      "epoch": 0.17,
      "learning_rate": 4.9857917669201445e-05,
      "loss": 0.0777,
      "step": 35
    },
    {
      "epoch": 0.19,
      "learning_rate": 4.981443720188455e-05,
      "loss": 0.1283,
      "step": 40
    },
    {
      "epoch": 0.22,
      "learning_rate": 4.97651852622661e-05,
      "loss": 0.0895,
      "step": 45
    },
    {
      "epoch": 0.24,
      "learning_rate": 4.9710173305621214e-05,
      "loss": 0.0873,
      "step": 50
    },
    {
      "epoch": 0.24,
      "eval_loss": 0.07891710102558136,
      "eval_runtime": 0.1169,
      "eval_samples_per_second": 3515.748,
      "eval_steps_per_second": 111.204,
      "step": 50
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.964941412692007e-05,
      "loss": 0.0729,
      "step": 55
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.9582921857851984e-05,
      "loss": 0.0668,
      "step": 60
    },
    {
      "epoch": 0.32,
      "learning_rate": 4.951071196353857e-05,
      "loss": 0.0755,
      "step": 65
    },
    {
      "epoch": 0.34,
      "learning_rate": 4.9432801238936795e-05,
      "loss": 0.1041,
      "step": 70
    },
    {
      "epoch": 0.36,
      "learning_rate": 4.9349207804932704e-05,
      "loss": 0.0446,
      "step": 75
    },
    {
      "epoch": 0.36,
      "eval_loss": 0.0668846070766449,
      "eval_runtime": 0.1152,
      "eval_samples_per_second": 3569.236,
      "eval_steps_per_second": 112.896,
      "step": 75
    },
    {
      "epoch": 0.39,
      "learning_rate": 4.925995110412681e-05,
      "loss": 0.0687,
      "step": 80
    },
    {
      "epoch": 0.41,
      "learning_rate": 4.916505189631196e-05,
      "loss": 0.0395,
      "step": 85
    },
    {
      "epoch": 0.44,
      "learning_rate": 4.906453225364499e-05,
      "loss": 0.0801,
      "step": 90
    },
    {
      "epoch": 0.46,
      "learning_rate": 4.8958415555513e-05,
      "loss": 0.0977,
      "step": 95
    },
    {
      "epoch": 0.49,
      "learning_rate": 4.884672648309572e-05,
      "loss": 0.0902,
      "step": 100
    },
    {
      "epoch": 0.49,
      "eval_loss": 0.06855440139770508,
      "eval_runtime": 0.1158,
      "eval_samples_per_second": 3547.794,
      "eval_steps_per_second": 112.217,
      "step": 100
    },
    {
      "epoch": 0.51,
      "learning_rate": 4.8729491013624956e-05,
      "loss": 0.0672,
      "step": 105
    },
    {
      "epoch": 0.53,
      "learning_rate": 4.860673641434275e-05,
      "loss": 0.0433,
      "step": 110
    },
    {
      "epoch": 0.56,
      "learning_rate": 4.847849123615934e-05,
      "loss": 0.0619,
      "step": 115
    },
    {
      "epoch": 0.58,
      "learning_rate": 4.834478530701271e-05,
      "loss": 0.0422,
      "step": 120
    },
    {
      "epoch": 0.61,
      "learning_rate": 4.820564972493101e-05,
      "loss": 0.1589,
      "step": 125
    },
    {
      "epoch": 0.61,
      "eval_loss": 0.15087510645389557,
      "eval_runtime": 0.1202,
      "eval_samples_per_second": 3418.754,
      "eval_steps_per_second": 108.136,
      "step": 125
    },
    {
      "epoch": 0.63,
      "learning_rate": 4.8061116850799625e-05,
      "loss": 0.147,
      "step": 130
    },
    {
      "epoch": 0.66,
      "learning_rate": 4.79112203008345e-05,
      "loss": 0.1062,
      "step": 135
    },
    {
      "epoch": 0.68,
      "learning_rate": 4.775599493876354e-05,
      "loss": 0.052,
      "step": 140
    },
    {
      "epoch": 0.7,
      "learning_rate": 4.759547686771774e-05,
      "loss": 0.0462,
      "step": 145
    },
    {
      "epoch": 0.73,
      "learning_rate": 4.742970342183424e-05,
      "loss": 0.1183,
      "step": 150
    },
    {
      "epoch": 0.73,
      "eval_loss": 0.0662284642457962,
      "eval_runtime": 0.1095,
      "eval_samples_per_second": 3755.038,
      "eval_steps_per_second": 118.772,
      "step": 150
    },
    {
      "epoch": 0.75,
      "learning_rate": 4.7258713157572866e-05,
      "loss": 0.0695,
      "step": 155
    },
    {
      "epoch": 0.78,
      "learning_rate": 4.70825458447485e-05,
      "loss": 0.049,
      "step": 160
    },
    {
      "epoch": 0.8,
      "learning_rate": 4.690124245728121e-05,
      "loss": 0.0352,
      "step": 165
    },
    {
      "epoch": 0.83,
      "learning_rate": 4.671484516366633e-05,
      "loss": 0.1032,
      "step": 170
    },
    {
      "epoch": 0.85,
      "learning_rate": 4.652339731716663e-05,
      "loss": 0.0627,
      "step": 175
    },
    {
      "epoch": 0.85,
      "eval_loss": 0.05767974629998207,
      "eval_runtime": 0.1141,
      "eval_samples_per_second": 3600.998,
      "eval_steps_per_second": 113.9,
      "step": 175
    },
    {
      "epoch": 0.87,
      "learning_rate": 4.6326943445729075e-05,
      "loss": 0.0352,
      "step": 180
    },
    {
      "epoch": 0.9,
      "learning_rate": 4.612552924162822e-05,
      "loss": 0.0294,
      "step": 185
    },
    {
      "epoch": 0.92,
      "learning_rate": 4.591920155083887e-05,
      "loss": 0.0485,
      "step": 190
    },
    {
      "epoch": 0.95,
      "learning_rate": 4.570800836214041e-05,
      "loss": 0.0332,
      "step": 195
    },
    {
      "epoch": 0.97,
      "learning_rate": 4.5491998795955314e-05,
      "loss": 0.0741,
      "step": 200
    },
    {
      "epoch": 0.97,
      "eval_loss": 0.06550680845975876,
      "eval_runtime": 0.1437,
      "eval_samples_per_second": 2859.8,
      "eval_steps_per_second": 90.456,
      "step": 200
    },
    {
      "epoch": 1.0,
      "learning_rate": 4.527122309292447e-05,
      "loss": 0.0169,
      "step": 205
    },
    {
      "epoch": 1.02,
      "learning_rate": 4.504573260222192e-05,
      "loss": 0.0181,
      "step": 210
    },
    {
      "epoch": 1.04,
      "learning_rate": 4.481557976961184e-05,
      "loss": 0.0415,
      "step": 215
    },
    {
      "epoch": 1.07,
      "learning_rate": 4.458081812525039e-05,
      "loss": 0.0272,
      "step": 220
    },
    {
      "epoch": 1.09,
      "learning_rate": 4.43415022712354e-05,
      "loss": 0.0475,
      "step": 225
    },
    {
      "epoch": 1.09,
      "eval_loss": 0.09047000855207443,
      "eval_runtime": 0.1222,
      "eval_samples_per_second": 3364.592,
      "eval_steps_per_second": 106.423,
      "step": 225
    },
    {
      "epoch": 1.12,
      "learning_rate": 4.4097687868906754e-05,
      "loss": 0.1496,
      "step": 230
    },
    {
      "epoch": 1.14,
      "learning_rate": 4.384943162590033e-05,
      "loss": 0.0454,
      "step": 235
    },
    {
      "epoch": 1.17,
      "learning_rate": 4.359679128295869e-05,
      "loss": 0.1215,
      "step": 240
    },
    {
      "epoch": 1.19,
      "learning_rate": 4.333982560050143e-05,
      "loss": 0.1333,
      "step": 245
    },
    {
      "epoch": 1.21,
      "learning_rate": 4.307859434495832e-05,
      "loss": 0.0372,
      "step": 250
    },
    {
      "epoch": 1.21,
      "eval_loss": 0.06259685754776001,
      "eval_runtime": 0.1186,
      "eval_samples_per_second": 3465.468,
      "eval_steps_per_second": 109.613,
      "step": 250
    },
    {
      "epoch": 1.24,
      "learning_rate": 4.281315827486857e-05,
      "loss": 0.0586,
      "step": 255
    },
    {
      "epoch": 1.26,
      "learning_rate": 4.2543579126749264e-05,
      "loss": 0.0252,
      "step": 260
    },
    {
      "epoch": 1.29,
      "learning_rate": 4.2269919600736385e-05,
      "loss": 0.0412,
      "step": 265
    },
    {
      "epoch": 1.31,
      "learning_rate": 4.199224334600163e-05,
      "loss": 0.0224,
      "step": 270
    },
    {
      "epoch": 1.33,
      "learning_rate": 4.17106149459486e-05,
      "loss": 0.0628,
      "step": 275
    },
    {
      "epoch": 1.33,
      "eval_loss": 0.07162802666425705,
      "eval_runtime": 0.1208,
      "eval_samples_per_second": 3401.425,
      "eval_steps_per_second": 107.588,
      "step": 275
    },
    {
      "epoch": 1.36,
      "learning_rate": 4.142509990319165e-05,
      "loss": 0.031,
      "step": 280
    },
    {
      "epoch": 1.38,
      "learning_rate": 4.1135764624320936e-05,
      "loss": 0.0918,
      "step": 285
    },
    {
      "epoch": 1.41,
      "learning_rate": 4.084267640445719e-05,
      "loss": 0.0433,
      "step": 290
    },
    {
      "epoch": 1.43,
      "learning_rate": 4.054590341159992e-05,
      "loss": 0.0525,
      "step": 295
    },
    {
      "epoch": 1.46,
      "learning_rate": 4.024551467077252e-05,
      "loss": 0.003,
      "step": 300
    },
    {
      "epoch": 1.46,
      "eval_loss": 0.05948002263903618,
      "eval_runtime": 0.122,
      "eval_samples_per_second": 3369.281,
      "eval_steps_per_second": 106.571,
      "step": 300
    },
    {
      "epoch": 1.48,
      "learning_rate": 3.9941580047968034e-05,
      "loss": 0.043,
      "step": 305
    },
    {
      "epoch": 1.5,
      "learning_rate": 3.963417023389942e-05,
      "loss": 0.0389,
      "step": 310
    },
    {
      "epoch": 1.53,
      "learning_rate": 3.932335672755787e-05,
      "loss": 0.0483,
      "step": 315
    },
    {
      "epoch": 1.55,
      "learning_rate": 3.900921181958323e-05,
      "loss": 0.1083,
      "step": 320
    },
    {
      "epoch": 1.58,
      "learning_rate": 3.869180857545026e-05,
      "loss": 0.0575,
      "step": 325
    },
    {
      "epoch": 1.58,
      "eval_loss": 0.07884568721055984,
      "eval_runtime": 0.1455,
      "eval_samples_per_second": 2825.706,
      "eval_steps_per_second": 89.378,
      "step": 325
    },
    {
      "epoch": 1.6,
      "learning_rate": 3.837122081847468e-05,
      "loss": 0.0458,
      "step": 330
    },
    {
      "epoch": 1.63,
      "learning_rate": 3.8047523112642955e-05,
      "loss": 0.0563,
      "step": 335
    },
    {
      "epoch": 1.65,
      "learning_rate": 3.772079074526981e-05,
      "loss": 0.0832,
      "step": 340
    },
    {
      "epoch": 1.67,
      "learning_rate": 3.73910997094875e-05,
      "loss": 0.0506,
      "step": 345
    },
    {
      "epoch": 1.7,
      "learning_rate": 3.705852668657096e-05,
      "loss": 0.0576,
      "step": 350
    },
    {
      "epoch": 1.7,
      "eval_loss": 0.04254939407110214,
      "eval_runtime": 0.1497,
      "eval_samples_per_second": 2746.393,
      "eval_steps_per_second": 86.869,
      "step": 350
    },
    {
      "epoch": 1.72,
      "learning_rate": 3.672314902810282e-05,
      "loss": 0.0546,
      "step": 355
    },
    {
      "epoch": 1.75,
      "learning_rate": 3.638504473798262e-05,
      "loss": 0.0272,
      "step": 360
    },
    {
      "epoch": 1.77,
      "learning_rate": 3.6044292454284215e-05,
      "loss": 0.0329,
      "step": 365
    },
    {
      "epoch": 1.8,
      "learning_rate": 3.570097143096573e-05,
      "loss": 0.0344,
      "step": 370
    },
    {
      "epoch": 1.82,
      "learning_rate": 3.5355161519436284e-05,
      "loss": 0.0121,
      "step": 375
    },
    {
      "epoch": 1.82,
      "eval_loss": 0.04299682751297951,
      "eval_runtime": 0.1175,
      "eval_samples_per_second": 3498.034,
      "eval_steps_per_second": 110.643,
      "step": 375
    },
    {
      "epoch": 1.84,
      "learning_rate": 3.500694314998367e-05,
      "loss": 0.0343,
      "step": 380
    },
    {
      "epoch": 1.87,
      "learning_rate": 3.465639731306753e-05,
      "loss": 0.0673,
      "step": 385
    },
    {
      "epoch": 1.89,
      "learning_rate": 3.430360554048208e-05,
      "loss": 0.044,
      "step": 390
    },
    {
      "epoch": 1.92,
      "learning_rate": 3.3948649886393114e-05,
      "loss": 0.044,
      "step": 395
    },
    {
      "epoch": 1.94,
      "learning_rate": 3.359161290825332e-05,
      "loss": 0.0488,
      "step": 400
    },
    {
      "epoch": 1.94,
      "eval_loss": 0.04617065563797951,
      "eval_runtime": 0.1182,
      "eval_samples_per_second": 3477.955,
      "eval_steps_per_second": 110.008,
      "step": 400
    },
    {
      "epoch": 1.97,
      "learning_rate": 3.323257764760073e-05,
      "loss": 0.0178,
      "step": 405
    },
    {
      "epoch": 1.99,
      "learning_rate": 3.2871627610744385e-05,
      "loss": 0.0255,
      "step": 410
    },
    {
      "epoch": 2.01,
      "learning_rate": 3.250884674934208e-05,
      "loss": 0.0377,
      "step": 415
    },
    {
      "epoch": 2.04,
      "learning_rate": 3.214431944087437e-05,
      "loss": 0.0309,
      "step": 420
    },
    {
      "epoch": 2.06,
      "learning_rate": 3.17781304690197e-05,
      "loss": 0.0667,
      "step": 425
    },
    {
      "epoch": 2.06,
      "eval_loss": 0.07217594981193542,
      "eval_runtime": 0.1187,
      "eval_samples_per_second": 3462.503,
      "eval_steps_per_second": 109.52,
      "step": 425
    },
    {
      "epoch": 2.09,
      "learning_rate": 3.14103650039349e-05,
      "loss": 0.0339,
      "step": 430
    },
    {
      "epoch": 2.11,
      "learning_rate": 3.104110858244588e-05,
      "loss": 0.0068,
      "step": 435
    },
    {
      "epoch": 2.14,
      "learning_rate": 3.0670447088153055e-05,
      "loss": 0.0522,
      "step": 440
    },
    {
      "epoch": 2.16,
      "learning_rate": 3.029846673145604e-05,
      "loss": 0.0068,
      "step": 445
    },
    {
      "epoch": 2.18,
      "learning_rate": 2.99252540295024e-05,
      "loss": 0.0514,
      "step": 450
    },
    {
      "epoch": 2.18,
      "eval_loss": 0.05946213752031326,
      "eval_runtime": 0.1357,
      "eval_samples_per_second": 3027.628,
      "eval_steps_per_second": 95.764,
      "step": 450
    },
    {
      "epoch": 2.21,
      "learning_rate": 2.955089578606506e-05,
      "loss": 0.022,
      "step": 455
    },
    {
      "epoch": 2.23,
      "learning_rate": 2.9175479071352996e-05,
      "loss": 0.058,
      "step": 460
    },
    {
      "epoch": 2.26,
      "learning_rate": 2.8799091201760008e-05,
      "loss": 0.0262,
      "step": 465
    },
    {
      "epoch": 2.28,
      "learning_rate": 2.8421819719556205e-05,
      "loss": 0.0108,
      "step": 470
    },
    {
      "epoch": 2.31,
      "learning_rate": 2.804375237252694e-05,
      "loss": 0.0286,
      "step": 475
    },
    {
      "epoch": 2.31,
      "eval_loss": 0.04657759144902229,
      "eval_runtime": 0.1207,
      "eval_samples_per_second": 3405.261,
      "eval_steps_per_second": 107.709,
      "step": 475
    },
    {
      "epoch": 2.33,
      "learning_rate": 2.7664977093564022e-05,
      "loss": 0.0122,
      "step": 480
    },
    {
      "epoch": 2.35,
      "learning_rate": 2.728558198021378e-05,
      "loss": 0.0286,
      "step": 485
    },
    {
      "epoch": 2.38,
      "learning_rate": 2.6905655274186892e-05,
      "loss": 0.0284,
      "step": 490
    },
    {
      "epoch": 2.4,
      "learning_rate": 2.6525285340834644e-05,
      "loss": 0.0223,
      "step": 495
    },
    {
      "epoch": 2.43,
      "learning_rate": 2.6144560648596466e-05,
      "loss": 0.0349,
      "step": 500
    },
    {
      "epoch": 2.43,
      "eval_loss": 0.05097094178199768,
      "eval_runtime": 0.1152,
      "eval_samples_per_second": 3566.29,
      "eval_steps_per_second": 112.802,
      "step": 500
    },
    {
      "epoch": 2.45,
      "learning_rate": 2.576356974842349e-05,
      "loss": 0.031,
      "step": 505
    },
    {
      "epoch": 2.48,
      "learning_rate": 2.538240125318287e-05,
      "loss": 0.0534,
      "step": 510
    },
    {
      "epoch": 2.5,
      "learning_rate": 2.5001143817047762e-05,
      "loss": 0.04,
      "step": 515
    },
    {
      "epoch": 2.52,
      "learning_rate": 2.4619886114877678e-05,
      "loss": 0.0463,
      "step": 520
    },
    {
      "epoch": 2.55,
      "learning_rate": 2.4238716821593986e-05,
      "loss": 0.0231,
      "step": 525
    },
    {
      "epoch": 2.55,
      "eval_loss": 0.04907093197107315,
      "eval_runtime": 0.1367,
      "eval_samples_per_second": 3007.163,
      "eval_steps_per_second": 95.117,
      "step": 525
    },
    {
      "epoch": 2.57,
      "learning_rate": 2.3857724591555443e-05,
      "loss": 0.0292,
      "step": 530
    },
    {
      "epoch": 2.6,
      "learning_rate": 2.347699803793854e-05,
      "loss": 0.0527,
      "step": 535
    },
    {
      "epoch": 2.62,
      "learning_rate": 2.309662571212731e-05,
      "loss": 0.0424,
      "step": 540
    },
    {
      "epoch": 2.65,
      "learning_rate": 2.2716696083117664e-05,
      "loss": 0.0363,
      "step": 545
    },
    {
      "epoch": 2.67,
      "learning_rate": 2.233729751694071e-05,
      "loss": 0.0301,
      "step": 550
    },
    {
      "epoch": 2.67,
      "eval_loss": 0.04202725738286972,
      "eval_runtime": 0.1194,
      "eval_samples_per_second": 3441.042,
      "eval_steps_per_second": 108.841,
      "step": 550
    },
    {
      "epoch": 2.69,
      "learning_rate": 2.1958518256110213e-05,
      "loss": 0.0638,
      "step": 555
    },
    {
      "epoch": 2.72,
      "learning_rate": 2.1580446399098604e-05,
      "loss": 0.0684,
      "step": 560
    },
    {
      "epoch": 2.74,
      "learning_rate": 2.1203169879846667e-05,
      "loss": 0.0249,
      "step": 565
    },
    {
      "epoch": 2.77,
      "learning_rate": 2.0826776447311296e-05,
      "loss": 0.0454,
      "step": 570
    },
    {
      "epoch": 2.79,
      "learning_rate": 2.0451353645056464e-05,
      "loss": 0.0617,
      "step": 575
    },
    {
      "epoch": 2.79,
      "eval_loss": 0.03725917637348175,
      "eval_runtime": 0.1154,
      "eval_samples_per_second": 3562.318,
      "eval_steps_per_second": 112.677,
      "step": 575
    },
    {
      "epoch": 2.82,
      "learning_rate": 2.0076988790891832e-05,
      "loss": 0.0323,
      "step": 580
    },
    {
      "epoch": 2.84,
      "learning_rate": 1.970376895656394e-05,
      "loss": 0.0442,
      "step": 585
    },
    {
      "epoch": 2.86,
      "learning_rate": 1.9331780947504582e-05,
      "loss": 0.0052,
      "step": 590
    },
    {
      "epoch": 2.89,
      "learning_rate": 1.8961111282641173e-05,
      "loss": 0.0171,
      "step": 595
    },
    {
      "epoch": 2.91,
      "learning_rate": 1.859184617427367e-05,
      "loss": 0.0305,
      "step": 600
    },
    {
      "epoch": 2.91,
      "eval_loss": 0.054718561470508575,
      "eval_runtime": 0.144,
      "eval_samples_per_second": 2853.575,
      "eval_steps_per_second": 90.259,
      "step": 600
    },
    {
      "epoch": 2.94,
      "learning_rate": 1.822407150802294e-05,
      "loss": 0.0405,
      "step": 605
    },
    {
      "epoch": 2.96,
      "learning_rate": 1.7857872822854935e-05,
      "loss": 0.0273,
      "step": 610
    },
    {
      "epoch": 2.99,
      "learning_rate": 1.7493335291185675e-05,
      "loss": 0.0324,
      "step": 615
    },
    {
      "epoch": 3.01,
      "learning_rate": 1.7130543699071327e-05,
      "loss": 0.0459,
      "step": 620
    },
    {
      "epoch": 3.03,
      "learning_rate": 1.67695824264883e-05,
      "loss": 0.0365,
      "step": 625
    },
    {
      "epoch": 3.03,
      "eval_loss": 0.04022263363003731,
      "eval_runtime": 0.1169,
      "eval_samples_per_second": 3516.3,
      "eval_steps_per_second": 111.221,
      "step": 625
    },
    {
      "epoch": 3.06,
      "learning_rate": 1.6410535427707634e-05,
      "loss": 0.0307,
      "step": 630
    },
    {
      "epoch": 3.08,
      "learning_rate": 1.6053486211768503e-05,
      "loss": 0.0316,
      "step": 635
    },
    {
      "epoch": 3.11,
      "learning_rate": 1.5698517823055242e-05,
      "loss": 0.0236,
      "step": 640
    },
    {
      "epoch": 3.13,
      "learning_rate": 1.5345712821982443e-05,
      "loss": 0.0187,
      "step": 645
    },
    {
      "epoch": 3.16,
      "learning_rate": 1.4995153265792642e-05,
      "loss": 0.0282,
      "step": 650
    },
    {
      "epoch": 3.16,
      "eval_loss": 0.04607125371694565,
      "eval_runtime": 0.1202,
      "eval_samples_per_second": 3419.582,
      "eval_steps_per_second": 108.162,
      "step": 650
    },
    {
      "epoch": 3.18,
      "learning_rate": 1.4646920689470967e-05,
      "loss": 0.013,
      "step": 655
    },
    {
      "epoch": 3.2,
      "learning_rate": 1.4301096086781363e-05,
      "loss": 0.0035,
      "step": 660
    },
    {
      "epoch": 3.23,
      "learning_rate": 1.3957759891428568e-05,
      "loss": 0.0043,
      "step": 665
    },
    {
      "epoch": 3.25,
      "learning_rate": 1.3616991958350494e-05,
      "loss": 0.0072,
      "step": 670
    },
    {
      "epoch": 3.28,
      "learning_rate": 1.3278871545145084e-05,
      "loss": 0.0391,
      "step": 675
    },
    {
      "epoch": 3.28,
      "eval_loss": 0.04361347481608391,
      "eval_runtime": 0.1163,
      "eval_samples_per_second": 3535.033,
      "eval_steps_per_second": 111.814,
      "step": 675
    },
    {
      "epoch": 3.3,
      "learning_rate": 1.294347729363618e-05,
      "loss": 0.0266,
      "step": 680
    },
    {
      "epoch": 3.33,
      "learning_rate": 1.2610887211582612e-05,
      "loss": 0.0256,
      "step": 685
    },
    {
      "epoch": 3.35,
      "learning_rate": 1.2281178654534734e-05,
      "loss": 0.0311,
      "step": 690
    },
    {
      "epoch": 3.37,
      "learning_rate": 1.1954428307842646e-05,
      "loss": 0.0054,
      "step": 695
    },
    {
      "epoch": 3.4,
      "learning_rate": 1.1630712168820351e-05,
      "loss": 0.0282,
      "step": 700
    },
    {
      "epoch": 3.4,
      "eval_loss": 0.04334261640906334,
      "eval_runtime": 0.1164,
      "eval_samples_per_second": 3530.458,
      "eval_steps_per_second": 111.669,
      "step": 700
    },
    {
      "epoch": 3.42,
      "learning_rate": 1.1310105529069844e-05,
      "loss": 0.0409,
      "step": 705
    },
    {
      "epoch": 3.45,
      "learning_rate": 1.0992682956969458e-05,
      "loss": 0.0331,
      "step": 710
    },
    {
      "epoch": 3.47,
      "learning_rate": 1.0678518280330296e-05,
      "loss": 0.0116,
      "step": 715
    },
    {
      "epoch": 3.5,
      "learning_rate": 1.0367684569225028e-05,
      "loss": 0.0037,
      "step": 720
    },
    {
      "epoch": 3.52,
      "learning_rate": 1.0060254118992837e-05,
      "loss": 0.0465,
      "step": 725
    },
    {
      "epoch": 3.52,
      "eval_loss": 0.048130132257938385,
      "eval_runtime": 0.1194,
      "eval_samples_per_second": 3441.283,
      "eval_steps_per_second": 108.848,
      "step": 725
    },
    {
      "epoch": 3.54,
      "learning_rate": 9.756298433424587e-06,
      "loss": 0.1006,
      "step": 730
    },
    {
      "epoch": 3.57,
      "learning_rate": 9.455888208132102e-06,
      "loss": 0.0071,
      "step": 735
    },
    {
      "epoch": 3.59,
      "learning_rate": 9.159093314105405e-06,
      "loss": 0.0139,
      "step": 740
    },
    {
      "epoch": 3.62,
      "learning_rate": 8.865982781461791e-06,
      "loss": 0.004,
      "step": 745
    },
    {
      "epoch": 3.64,
      "learning_rate": 8.576624783390367e-06,
      "loss": 0.0557,
      "step": 750
    },
    {
      "epoch": 3.64,
      "eval_loss": 0.04551706090569496,
      "eval_runtime": 0.1186,
      "eval_samples_per_second": 3464.82,
      "eval_steps_per_second": 109.593,
      "step": 750
    },
    {
      "epoch": 3.67,
      "learning_rate": 8.291086620296052e-06,
      "loss": 0.0347,
      "step": 755
    },
    {
      "epoch": 3.69,
      "learning_rate": 8.009434704146424e-06,
      "loss": 0.0413,
      "step": 760
    },
    {
      "epoch": 3.71,
      "learning_rate": 7.731734543025345e-06,
      "loss": 0.0147,
      "step": 765
    },
    {
      "epoch": 3.74,
      "learning_rate": 7.458050725896673e-06,
      "loss": 0.0502,
      "step": 770
    },
    {
      "epoch": 3.76,
      "learning_rate": 7.188446907581894e-06,
      "loss": 0.058,
      "step": 775
    },
    {
      "epoch": 3.76,
      "eval_loss": 0.04955735430121422,
      "eval_runtime": 0.1466,
      "eval_samples_per_second": 2802.9,
      "eval_steps_per_second": 88.656,
      "step": 775
    },
    {
      "epoch": 3.79,
      "learning_rate": 6.922985793954881e-06,
      "loss": 0.0127,
      "step": 780
    },
    {
      "epoch": 3.81,
      "learning_rate": 6.661729127357494e-06,
      "loss": 0.0335,
      "step": 785
    },
    {
      "epoch": 3.83,
      "learning_rate": 6.404737672239173e-06,
      "loss": 0.0271,
      "step": 790
    },
    {
      "epoch": 3.86,
      "learning_rate": 6.1520712010240455e-06,
      "loss": 0.0512,
      "step": 795
    },
    {
      "epoch": 3.88,
      "learning_rate": 5.9037884802087325e-06,
      "loss": 0.0057,
      "step": 800
    },
    {
      "epoch": 3.88,
      "eval_loss": 0.051423329859972,
      "eval_runtime": 0.1163,
      "eval_samples_per_second": 3533.425,
      "eval_steps_per_second": 111.763,
      "step": 800
    },
    {
      "epoch": 3.91,
      "learning_rate": 5.659947256694156e-06,
      "loss": 0.0029,
      "step": 805
    },
    {
      "epoch": 3.93,
      "learning_rate": 5.420604244354408e-06,
      "loss": 0.0095,
      "step": 810
    },
    {
      "epoch": 3.96,
      "learning_rate": 5.185815110845996e-06,
      "loss": 0.0491,
      "step": 815
    },
    {
      "epoch": 3.98,
      "learning_rate": 4.9556344646603165e-06,
      "loss": 0.0251,
      "step": 820
    },
    {
      "epoch": 4.0,
      "learning_rate": 4.73011584242257e-06,
      "loss": 0.032,
      "step": 825
    },
    {
      "epoch": 4.0,
      "eval_loss": 0.05062270909547806,
      "eval_runtime": 0.1194,
      "eval_samples_per_second": 3442.19,
      "eval_steps_per_second": 108.877,
      "step": 825
    },
    {
      "epoch": 4.03,
      "learning_rate": 4.509311696439903e-06,
      "loss": 0.0354,
      "step": 830
    },
    {
      "epoch": 4.05,
      "learning_rate": 4.293273382501775e-06,
      "loss": 0.0737,
      "step": 835
    },
    {
      "epoch": 4.08,
      "learning_rate": 4.082051147935373e-06,
      "loss": 0.0198,
      "step": 840
    },
    {
      "epoch": 4.1,
      "learning_rate": 3.875694119918805e-06,
      "loss": 0.0266,
      "step": 845
    },
    {
      "epoch": 4.13,
      "learning_rate": 3.6742502940548734e-06,
      "loss": 0.0056,
      "step": 850
    },
    {
      "epoch": 4.13,
      "eval_loss": 0.048265255987644196,
      "eval_runtime": 0.1135,
      "eval_samples_per_second": 3619.583,
      "eval_steps_per_second": 114.488,
      "step": 850
    },
    {
      "epoch": 4.15,
      "learning_rate": 3.477766523207965e-06,
      "loss": 0.0017,
      "step": 855
    },
    {
      "epoch": 4.17,
      "learning_rate": 3.286288506606805e-06,
      "loss": 0.0009,
      "step": 860
    },
    {
      "epoch": 4.2,
      "learning_rate": 3.0998607792154495e-06,
      "loss": 0.0153,
      "step": 865
    },
    {
      "epoch": 4.22,
      "learning_rate": 2.91852670137516e-06,
      "loss": 0.0232,
      "step": 870
    },
    {
      "epoch": 4.25,
      "learning_rate": 2.7423284487193697e-06,
      "loss": 0.0104,
      "step": 875
    },
    {
      "epoch": 4.25,
      "eval_loss": 0.048452552407979965,
      "eval_runtime": 0.1158,
      "eval_samples_per_second": 3550.622,
      "eval_steps_per_second": 112.307,
      "step": 875
    },
    {
      "epoch": 4.27,
      "learning_rate": 2.571307002364301e-06,
      "loss": 0.0146,
      "step": 880
    },
    {
      "epoch": 4.3,
      "learning_rate": 2.4055021393773277e-06,
      "loss": 0.062,
      "step": 885
    },
    {
      "epoch": 4.32,
      "learning_rate": 2.2449524235254484e-06,
      "loss": 0.0415,
      "step": 890
    },
    {
      "epoch": 4.34,
      "learning_rate": 2.089695196305888e-06,
      "loss": 0.0037,
      "step": 895
    },
    {
      "epoch": 4.37,
      "learning_rate": 1.939766568261037e-06,
      "loss": 0.0353,
      "step": 900
    },
    {
      "epoch": 4.37,
      "eval_loss": 0.04834846034646034,
      "eval_runtime": 0.1173,
      "eval_samples_per_second": 3503.757,
      "eval_steps_per_second": 110.824,
      "step": 900
    },
    {
      "epoch": 4.39,
      "learning_rate": 1.7952014105796666e-06,
      "loss": 0.0316,
      "step": 905
    },
    {
      "epoch": 4.42,
      "learning_rate": 1.656033346986416e-06,
      "loss": 0.0238,
      "step": 910
    },
    {
      "epoch": 4.44,
      "learning_rate": 1.5222947459213793e-06,
      "loss": 0.025,
      "step": 915
    },
    {
      "epoch": 4.47,
      "learning_rate": 1.3940167130117226e-06,
      "loss": 0.0076,
      "step": 920
    },
    {
      "epoch": 4.49,
      "learning_rate": 1.2712290838369363e-06,
      "loss": 0.0044,
      "step": 925
    },
    {
      "epoch": 4.49,
      "eval_loss": 0.0482080839574337,
      "eval_runtime": 0.1126,
      "eval_samples_per_second": 3651.3,
      "eval_steps_per_second": 115.491,
      "step": 925
    },
    {
      "epoch": 4.51,
      "learning_rate": 1.1539604169895497e-06,
      "loss": 0.0038,
      "step": 930
    },
    {
      "epoch": 4.54,
      "learning_rate": 1.0422379874328008e-06,
      "loss": 0.0038,
      "step": 935
    },
    {
      "epoch": 4.56,
      "learning_rate": 9.360877801568968e-07,
      "loss": 0.0463,
      "step": 940
    },
    {
      "epoch": 4.59,
      "learning_rate": 8.355344841352836e-07,
      "loss": 0.001,
      "step": 945
    },
    {
      "epoch": 4.61,
      "learning_rate": 7.406014865823513e-07,
      "loss": 0.0449,
      "step": 950
    },
    {
      "epoch": 4.61,
      "eval_loss": 0.04827665537595749,
      "eval_runtime": 0.1168,
      "eval_samples_per_second": 3519.71,
      "eval_steps_per_second": 111.329,
      "step": 950
    },
    {
      "epoch": 4.64,
      "learning_rate": 6.513108675139101e-07,
      "loss": 0.0138,
      "step": 955
    },
    {
      "epoch": 4.66,
      "learning_rate": 5.676833946117205e-07,
      "loss": 0.0407,
      "step": 960
    },
    {
      "epoch": 4.68,
      "learning_rate": 4.897385183932179e-07,
      "loss": 0.0392,
      "step": 965
    },
    {
      "epoch": 4.71,
      "learning_rate": 4.1749436768762084e-07,
      "loss": 0.0358,
      "step": 970
    },
    {
      "epoch": 4.73,
      "learning_rate": 3.509677454194282e-07,
      "loss": 0.0626,
      "step": 975
    },
    {
      "epoch": 4.73,
      "eval_loss": 0.04846416041254997,
      "eval_runtime": 0.116,
      "eval_samples_per_second": 3543.615,
      "eval_steps_per_second": 112.085,
      "step": 975
    },
    {
      "epoch": 4.76,
      "learning_rate": 2.901741247002987e-07,
      "loss": 0.0069,
      "step": 980
    },
    {
      "epoch": 4.78,
      "learning_rate": 2.3512764523025698e-07,
      "loss": 0.0697,
      "step": 985
    },
    {
      "epoch": 4.81,
      "learning_rate": 1.8584111000897464e-07,
      "loss": 0.03,
      "step": 990
    },
    {
      "epoch": 4.83,
      "learning_rate": 1.423259823580131e-07,
      "loss": 0.0007,
      "step": 995
    },
    {
      "epoch": 4.85,
      "learning_rate": 1.0459238325460363e-07,
      "loss": 0.0631,
      "step": 1000
    },
    {
      "epoch": 4.85,
      "eval_loss": 0.0483647957444191,
      "eval_runtime": 0.146,
      "eval_samples_per_second": 2815.962,
      "eval_steps_per_second": 89.069,
      "step": 1000
    },
    {
      "epoch": 4.88,
      "learning_rate": 7.264908897766098e-08,
      "loss": 0.0569,
      "step": 1005
    },
    {
      "epoch": 4.9,
      "learning_rate": 4.650352906655775e-08,
      "loss": 0.0304,
      "step": 1010
    },
    {
      "epoch": 4.93,
      "learning_rate": 2.616178459311225e-08,
      "loss": 0.0101,
      "step": 1015
    },
    {
      "epoch": 4.95,
      "learning_rate": 1.1628586747233772e-08,
      "loss": 0.0035,
      "step": 1020
    },
    {
      "epoch": 4.98,
      "learning_rate": 2.9073157365056002e-09,
      "loss": 0.0054,
      "step": 1025
    },
    {
      "epoch": 4.98,
      "eval_loss": 0.04826511815190315,
      "eval_runtime": 0.1121,
      "eval_samples_per_second": 3664.947,
      "eval_steps_per_second": 115.923,
      "step": 1025
    },
    {
      "epoch": 5.0,
      "learning_rate": 0.0,
      "loss": 0.003,
      "step": 1030
    }
  ],
  "logging_steps": 5,
  "max_steps": 1030,
  "num_train_epochs": 5,
  "save_steps": 500,
  "total_flos": 459260224208896.0,
  "trial_name": null,
  "trial_params": null
}