{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.0,
  "eval_steps": 500,
  "global_step": 5000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.02,
      "grad_norm": NaN,
      "learning_rate": 2.4000000000000003e-07,
      "loss": 1.3992,
      "step": 10
    },
    {
      "epoch": 0.04,
      "grad_norm": 1.1072826385498047,
      "learning_rate": 6.000000000000001e-07,
      "loss": 1.4497,
      "step": 20
    },
    {
      "epoch": 0.06,
      "grad_norm": 3.9411494731903076,
      "learning_rate": 1.0000000000000002e-06,
      "loss": 1.2599,
      "step": 30
    },
    {
      "epoch": 0.08,
      "grad_norm": 3.928187847137451,
      "learning_rate": 1.4000000000000001e-06,
      "loss": 1.4124,
      "step": 40
    },
    {
      "epoch": 0.1,
      "grad_norm": 0.6260775923728943,
      "learning_rate": 1.8000000000000001e-06,
      "loss": 1.3542,
      "step": 50
    },
    {
      "epoch": 0.12,
      "grad_norm": 1.5522129535675049,
      "learning_rate": 2.2e-06,
      "loss": 1.2123,
      "step": 60
    },
    {
      "epoch": 0.14,
      "grad_norm": 1.6981067657470703,
      "learning_rate": 2.6e-06,
      "loss": 1.2098,
      "step": 70
    },
    {
      "epoch": 0.16,
      "grad_norm": 5.6391496658325195,
      "learning_rate": 3e-06,
      "loss": 1.2226,
      "step": 80
    },
    {
      "epoch": 0.18,
      "grad_norm": 0.5151563286781311,
      "learning_rate": 3.3600000000000004e-06,
      "loss": 1.0624,
      "step": 90
    },
    {
      "epoch": 0.2,
      "grad_norm": 1.4428874254226685,
      "learning_rate": 3.7600000000000004e-06,
      "loss": 1.0969,
      "step": 100
    },
    {
      "epoch": 0.22,
      "grad_norm": 2.130734920501709,
      "learning_rate": 4.16e-06,
      "loss": 1.0879,
      "step": 110
    },
    {
      "epoch": 0.24,
      "grad_norm": 2.011439561843872,
      "learning_rate": 4.56e-06,
      "loss": 1.0195,
      "step": 120
    },
    {
      "epoch": 0.26,
      "grad_norm": 2.217247486114502,
      "learning_rate": 4.960000000000001e-06,
      "loss": 0.9765,
      "step": 130
    },
    {
      "epoch": 0.28,
      "grad_norm": 0.946567952632904,
      "learning_rate": 5.36e-06,
      "loss": 1.0205,
      "step": 140
    },
    {
      "epoch": 0.3,
      "grad_norm": 6.922210693359375,
      "learning_rate": 5.76e-06,
      "loss": 0.9517,
      "step": 150
    },
    {
      "epoch": 0.32,
      "grad_norm": 3.384115219116211,
      "learning_rate": 6.16e-06,
      "loss": 0.9324,
      "step": 160
    },
    {
      "epoch": 0.34,
      "grad_norm": 1.7061117887496948,
      "learning_rate": 6.560000000000001e-06,
      "loss": 0.804,
      "step": 170
    },
    {
      "epoch": 0.36,
      "grad_norm": 1.5616205930709839,
      "learning_rate": 6.96e-06,
      "loss": 0.7821,
      "step": 180
    },
    {
      "epoch": 0.38,
      "grad_norm": 1.8393518924713135,
      "learning_rate": 7.360000000000001e-06,
      "loss": 0.8086,
      "step": 190
    },
    {
      "epoch": 0.4,
      "grad_norm": 1.4879248142242432,
      "learning_rate": 7.76e-06,
      "loss": 0.7655,
      "step": 200
    },
    {
      "epoch": 0.42,
      "grad_norm": 0.6370295882225037,
      "learning_rate": 8.16e-06,
      "loss": 0.7508,
      "step": 210
    },
    {
      "epoch": 0.44,
      "grad_norm": 0.5269752144813538,
      "learning_rate": 8.560000000000001e-06,
      "loss": 0.7429,
      "step": 220
    },
    {
      "epoch": 0.46,
      "grad_norm": 1.2695356607437134,
      "learning_rate": 8.96e-06,
      "loss": 0.7502,
      "step": 230
    },
    {
      "epoch": 0.48,
      "grad_norm": 1.5492205619812012,
      "learning_rate": 9.360000000000002e-06,
      "loss": 0.7029,
      "step": 240
    },
    {
      "epoch": 0.5,
      "grad_norm": 1.7380893230438232,
      "learning_rate": 9.760000000000001e-06,
      "loss": 0.7324,
      "step": 250
    },
    {
      "epoch": 0.52,
      "grad_norm": 1.8911452293395996,
      "learning_rate": 1.0160000000000001e-05,
      "loss": 0.7521,
      "step": 260
    },
    {
      "epoch": 0.54,
      "grad_norm": 2.0408151149749756,
      "learning_rate": 1.056e-05,
      "loss": 0.698,
      "step": 270
    },
    {
      "epoch": 0.56,
      "grad_norm": 1.9015631675720215,
      "learning_rate": 1.0960000000000002e-05,
      "loss": 0.6859,
      "step": 280
    },
    {
      "epoch": 0.58,
      "grad_norm": 1.5284056663513184,
      "learning_rate": 1.136e-05,
      "loss": 0.6716,
      "step": 290
    },
    {
      "epoch": 0.6,
      "grad_norm": 1.3547126054763794,
      "learning_rate": 1.1760000000000001e-05,
      "loss": 0.6978,
      "step": 300
    },
    {
      "epoch": 0.62,
      "grad_norm": 1.4986441135406494,
      "learning_rate": 1.216e-05,
      "loss": 0.6584,
      "step": 310
    },
    {
      "epoch": 0.64,
      "grad_norm": 1.1250969171524048,
      "learning_rate": 1.2560000000000002e-05,
      "loss": 0.7188,
      "step": 320
    },
    {
      "epoch": 0.66,
      "grad_norm": 1.1186408996582031,
      "learning_rate": 1.2960000000000001e-05,
      "loss": 0.6687,
      "step": 330
    },
    {
      "epoch": 0.68,
      "grad_norm": 1.1250578165054321,
      "learning_rate": 1.3360000000000003e-05,
      "loss": 0.623,
      "step": 340
    },
    {
      "epoch": 0.7,
      "grad_norm": 0.8931149840354919,
      "learning_rate": 1.376e-05,
      "loss": 0.6795,
      "step": 350
    },
    {
      "epoch": 0.72,
      "grad_norm": 1.5052251815795898,
      "learning_rate": 1.416e-05,
      "loss": 0.6455,
      "step": 360
    },
    {
      "epoch": 0.74,
      "grad_norm": 1.8043763637542725,
      "learning_rate": 1.4560000000000001e-05,
      "loss": 0.6548,
      "step": 370
    },
    {
      "epoch": 0.76,
      "grad_norm": 1.7357759475708008,
      "learning_rate": 1.496e-05,
      "loss": 0.6508,
      "step": 380
    },
    {
      "epoch": 0.78,
      "grad_norm": 1.4080872535705566,
      "learning_rate": 1.5360000000000002e-05,
      "loss": 0.7105,
      "step": 390
    },
    {
      "epoch": 0.8,
      "grad_norm": 0.7395206689834595,
      "learning_rate": 1.576e-05,
      "loss": 0.6738,
      "step": 400
    },
    {
      "epoch": 0.82,
      "grad_norm": 1.4803540706634521,
      "learning_rate": 1.616e-05,
      "loss": 0.6741,
      "step": 410
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.9231945276260376,
      "learning_rate": 1.656e-05,
      "loss": 0.6385,
      "step": 420
    },
    {
      "epoch": 0.86,
      "grad_norm": 1.0000849962234497,
      "learning_rate": 1.696e-05,
      "loss": 0.6304,
      "step": 430
    },
    {
      "epoch": 0.88,
      "grad_norm": 1.8198318481445312,
      "learning_rate": 1.736e-05,
      "loss": 0.652,
      "step": 440
    },
    {
      "epoch": 0.9,
      "grad_norm": 0.8213591575622559,
      "learning_rate": 1.7760000000000003e-05,
      "loss": 0.6517,
      "step": 450
    },
    {
      "epoch": 0.92,
      "grad_norm": 2.0881271362304688,
      "learning_rate": 1.8160000000000002e-05,
      "loss": 0.7044,
      "step": 460
    },
    {
      "epoch": 0.94,
      "grad_norm": 2.583738088607788,
      "learning_rate": 1.8560000000000002e-05,
      "loss": 0.6801,
      "step": 470
    },
    {
      "epoch": 0.96,
      "grad_norm": 2.3651039600372314,
      "learning_rate": 1.896e-05,
      "loss": 0.6531,
      "step": 480
    },
    {
      "epoch": 0.98,
      "grad_norm": 1.174816608428955,
      "learning_rate": 1.936e-05,
      "loss": 0.6837,
      "step": 490
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.93290376663208,
      "learning_rate": 1.976e-05,
      "loss": 0.6659,
      "step": 500
    },
    {
      "epoch": 1.02,
      "grad_norm": 1.616199254989624,
      "learning_rate": 1.9999961008995607e-05,
      "loss": 0.6212,
      "step": 510
    },
    {
      "epoch": 1.04,
      "grad_norm": 1.4221971035003662,
      "learning_rate": 1.99995223636881e-05,
      "loss": 0.6705,
      "step": 520
    },
    {
      "epoch": 1.06,
      "grad_norm": 1.4205855131149292,
      "learning_rate": 1.9998596355767805e-05,
      "loss": 0.6346,
      "step": 530
    },
    {
      "epoch": 1.08,
      "grad_norm": 1.5640236139297485,
      "learning_rate": 1.999718303036705e-05,
      "loss": 0.6698,
      "step": 540
    },
    {
      "epoch": 1.1,
      "grad_norm": 1.3953174352645874,
      "learning_rate": 1.9995282456369313e-05,
      "loss": 0.5925,
      "step": 550
    },
    {
      "epoch": 1.12,
      "grad_norm": 1.3201889991760254,
      "learning_rate": 1.9992894726405894e-05,
      "loss": 0.6795,
      "step": 560
    },
    {
      "epoch": 1.1400000000000001,
      "grad_norm": 1.8795799016952515,
      "learning_rate": 1.9990019956851384e-05,
      "loss": 0.6096,
      "step": 570
    },
    {
      "epoch": 1.16,
      "grad_norm": 1.9675489664077759,
      "learning_rate": 1.998665828781799e-05,
      "loss": 0.5971,
      "step": 580
    },
    {
      "epoch": 1.18,
      "grad_norm": 1.0514065027236938,
      "learning_rate": 1.998280988314872e-05,
      "loss": 0.6055,
      "step": 590
    },
    {
      "epoch": 1.2,
      "grad_norm": 0.7312430143356323,
      "learning_rate": 1.9978474930409396e-05,
      "loss": 0.6304,
      "step": 600
    },
    {
      "epoch": 1.22,
      "grad_norm": 11.12009048461914,
      "learning_rate": 1.9973653640879486e-05,
      "loss": 0.6812,
      "step": 610
    },
    {
      "epoch": 1.24,
      "grad_norm": 2.484487771987915,
      "learning_rate": 1.9968346249541848e-05,
      "loss": 0.5842,
      "step": 620
    },
    {
      "epoch": 1.26,
      "grad_norm": 1.4185147285461426,
      "learning_rate": 1.996255301507125e-05,
      "loss": 0.6478,
      "step": 630
    },
    {
      "epoch": 1.28,
      "grad_norm": 0.939939022064209,
      "learning_rate": 1.995627421982176e-05,
      "loss": 0.6003,
      "step": 640
    },
    {
      "epoch": 1.3,
      "grad_norm": 2.303175687789917,
      "learning_rate": 1.9949510169813006e-05,
      "loss": 0.6087,
      "step": 650
    },
    {
      "epoch": 1.32,
      "grad_norm": 1.1472731828689575,
      "learning_rate": 1.9942261194715236e-05,
      "loss": 0.5905,
      "step": 660
    },
    {
      "epoch": 1.34,
      "grad_norm": 0.8142107129096985,
      "learning_rate": 1.9934527647833276e-05,
      "loss": 0.593,
      "step": 670
    },
    {
      "epoch": 1.3599999999999999,
      "grad_norm": 1.4084548950195312,
      "learning_rate": 1.992630990608929e-05,
      "loss": 0.6253,
      "step": 680
    },
    {
      "epoch": 1.38,
      "grad_norm": 1.6054160594940186,
      "learning_rate": 1.9917608370004417e-05,
      "loss": 0.6338,
      "step": 690
    },
    {
      "epoch": 1.4,
      "grad_norm": 2.442007541656494,
      "learning_rate": 1.9908423463679246e-05,
      "loss": 0.6148,
      "step": 700
    },
    {
      "epoch": 1.42,
      "grad_norm": 0.7885130047798157,
      "learning_rate": 1.989875563477316e-05,
      "loss": 0.6271,
      "step": 710
    },
    {
      "epoch": 1.44,
      "grad_norm": 2.8302054405212402,
      "learning_rate": 1.9888605354482494e-05,
      "loss": 0.638,
      "step": 720
    },
    {
      "epoch": 1.46,
      "grad_norm": 1.134475827217102,
      "learning_rate": 1.987797311751759e-05,
      "loss": 0.6304,
      "step": 730
    },
    {
      "epoch": 1.48,
      "grad_norm": 0.7951676249504089,
      "learning_rate": 1.986685944207868e-05,
      "loss": 0.5877,
      "step": 740
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.5261722803115845,
      "learning_rate": 1.985526486983063e-05,
      "loss": 0.5747,
      "step": 750
    },
    {
      "epoch": 1.52,
      "grad_norm": 4.633131980895996,
      "learning_rate": 1.9843189965876525e-05,
      "loss": 0.6514,
      "step": 760
    },
    {
      "epoch": 1.54,
      "grad_norm": 0.7688568830490112,
      "learning_rate": 1.9830635318730155e-05,
      "loss": 0.5879,
      "step": 770
    },
    {
      "epoch": 1.56,
      "grad_norm": 0.859425961971283,
      "learning_rate": 1.981760154028731e-05,
      "loss": 0.6152,
      "step": 780
    },
    {
      "epoch": 1.58,
      "grad_norm": 1.939563274383545,
      "learning_rate": 1.980408926579596e-05,
      "loss": 0.6342,
      "step": 790
    },
    {
      "epoch": 1.6,
      "grad_norm": 0.7612221837043762,
      "learning_rate": 1.97900991538253e-05,
      "loss": 0.6167,
      "step": 800
    },
    {
      "epoch": 1.62,
      "grad_norm": 2.2331180572509766,
      "learning_rate": 1.9775631886233655e-05,
      "loss": 0.5688,
      "step": 810
    },
    {
      "epoch": 1.6400000000000001,
      "grad_norm": 3.1707897186279297,
      "learning_rate": 1.9760688168135233e-05,
      "loss": 0.6023,
      "step": 820
    },
    {
      "epoch": 1.6600000000000001,
      "grad_norm": 1.5035152435302734,
      "learning_rate": 1.9745268727865774e-05,
      "loss": 0.5957,
      "step": 830
    },
    {
      "epoch": 1.6800000000000002,
      "grad_norm": 1.8150962591171265,
      "learning_rate": 1.972937431694704e-05,
      "loss": 0.5409,
      "step": 840
    },
    {
      "epoch": 1.7,
      "grad_norm": 1.5338727235794067,
      "learning_rate": 1.9713005710050203e-05,
      "loss": 0.6286,
      "step": 850
    },
    {
      "epoch": 1.72,
      "grad_norm": 1.0299500226974487,
      "learning_rate": 1.969616370495806e-05,
      "loss": 0.5936,
      "step": 860
    },
    {
      "epoch": 1.74,
      "grad_norm": 3.0667810440063477,
      "learning_rate": 1.967884912252619e-05,
      "loss": 0.6535,
      "step": 870
    },
    {
      "epoch": 1.76,
      "grad_norm": 0.9165984988212585,
      "learning_rate": 1.9661062806642903e-05,
      "loss": 0.5864,
      "step": 880
    },
    {
      "epoch": 1.78,
      "grad_norm": 1.398553729057312,
      "learning_rate": 1.964280562418815e-05,
      "loss": 0.6181,
      "step": 890
    },
    {
      "epoch": 1.8,
      "grad_norm": 1.1232646703720093,
      "learning_rate": 1.962407846499124e-05,
      "loss": 0.5736,
      "step": 900
    },
    {
      "epoch": 1.8199999999999998,
      "grad_norm": 1.905674695968628,
      "learning_rate": 1.96048822417875e-05,
      "loss": 0.5769,
      "step": 910
    },
    {
      "epoch": 1.8399999999999999,
      "grad_norm": 0.9369404911994934,
      "learning_rate": 1.958521789017376e-05,
      "loss": 0.6056,
      "step": 920
    },
    {
      "epoch": 1.8599999999999999,
      "grad_norm": 1.0187280178070068,
      "learning_rate": 1.956508636856278e-05,
      "loss": 0.6632,
      "step": 930
    },
    {
      "epoch": 1.88,
      "grad_norm": 1.4954912662506104,
      "learning_rate": 1.9546569379242446e-05,
      "loss": 0.5803,
      "step": 940
    },
    {
      "epoch": 1.9,
      "grad_norm": 1.1077364683151245,
      "learning_rate": 1.9525552956573244e-05,
      "loss": 0.6028,
      "step": 950
    },
    {
      "epoch": 1.92,
      "grad_norm": 1.777431607246399,
      "learning_rate": 1.9504072271891486e-05,
      "loss": 0.5932,
      "step": 960
    },
    {
      "epoch": 1.94,
      "grad_norm": 1.1748569011688232,
      "learning_rate": 1.9482128372135446e-05,
      "loss": 0.581,
      "step": 970
    },
    {
      "epoch": 1.96,
      "grad_norm": 1.5400365591049194,
      "learning_rate": 1.945972232681984e-05,
      "loss": 0.6207,
      "step": 980
    },
    {
      "epoch": 1.98,
      "grad_norm": 1.3064255714416504,
      "learning_rate": 1.9436855227983695e-05,
      "loss": 0.5576,
      "step": 990
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.6377707719802856,
      "learning_rate": 1.9413528190137158e-05,
      "loss": 0.6121,
      "step": 1000
    },
    {
      "epoch": 2.02,
      "grad_norm": 0.8060258030891418,
      "learning_rate": 1.938974235020714e-05,
      "loss": 0.5909,
      "step": 1010
    },
    {
      "epoch": 2.04,
      "grad_norm": 2.174511671066284,
      "learning_rate": 1.9365498867481926e-05,
      "loss": 0.5459,
      "step": 1020
    },
    {
      "epoch": 2.06,
      "grad_norm": 2.0883657932281494,
      "learning_rate": 1.9340798923554657e-05,
      "loss": 0.5781,
      "step": 1030
    },
    {
      "epoch": 2.08,
      "grad_norm": 1.4199424982070923,
      "learning_rate": 1.931564372226576e-05,
      "loss": 0.5284,
      "step": 1040
    },
    {
      "epoch": 2.1,
      "grad_norm": 1.2759448289871216,
      "learning_rate": 1.9290034489644247e-05,
      "loss": 0.5476,
      "step": 1050
    },
    {
      "epoch": 2.12,
      "grad_norm": 0.9871682524681091,
      "learning_rate": 1.9263972473847995e-05,
      "loss": 0.5386,
      "step": 1060
    },
    {
      "epoch": 2.14,
      "grad_norm": 1.325584053993225,
      "learning_rate": 1.923745894510288e-05,
      "loss": 0.5864,
      "step": 1070
    },
    {
      "epoch": 2.16,
      "grad_norm": 1.4707293510437012,
      "learning_rate": 1.9210495195640895e-05,
      "loss": 0.5413,
      "step": 1080
    },
    {
      "epoch": 2.18,
      "grad_norm": 1.222374439239502,
      "learning_rate": 1.918308253963715e-05,
      "loss": 0.5201,
      "step": 1090
    },
    {
      "epoch": 2.2,
      "grad_norm": 2.6065587997436523,
      "learning_rate": 1.9155222313145817e-05,
      "loss": 0.5658,
      "step": 1100
    },
    {
      "epoch": 2.22,
      "grad_norm": 1.2004817724227905,
      "learning_rate": 1.912691587403503e-05,
      "loss": 0.5578,
      "step": 1110
    },
    {
      "epoch": 2.24,
      "grad_norm": 2.0765957832336426,
      "learning_rate": 1.9098164601920702e-05,
      "loss": 0.4792,
      "step": 1120
    },
    {
      "epoch": 2.26,
      "grad_norm": 1.2682489156723022,
      "learning_rate": 1.906896989809927e-05,
      "loss": 0.6048,
      "step": 1130
    },
    {
      "epoch": 2.2800000000000002,
      "grad_norm": 3.5352697372436523,
      "learning_rate": 1.903933318547942e-05,
      "loss": 0.567,
      "step": 1140
    },
    {
      "epoch": 2.3,
      "grad_norm": 1.590265154838562,
      "learning_rate": 1.9009255908512704e-05,
      "loss": 0.5965,
      "step": 1150
    },
    {
      "epoch": 2.32,
      "grad_norm": 2.2210605144500732,
      "learning_rate": 1.897873953312317e-05,
      "loss": 0.5561,
      "step": 1160
    },
    {
      "epoch": 2.34,
      "grad_norm": 1.0093090534210205,
      "learning_rate": 1.8947785546635905e-05,
      "loss": 0.5529,
      "step": 1170
    },
    {
      "epoch": 2.36,
      "grad_norm": 1.523364543914795,
      "learning_rate": 1.8916395457704536e-05,
      "loss": 0.5818,
      "step": 1180
    },
    {
      "epoch": 2.38,
      "grad_norm": 0.9908381700515747,
      "learning_rate": 1.888457079623772e-05,
      "loss": 0.5558,
      "step": 1190
    },
    {
      "epoch": 2.4,
      "grad_norm": 1.499123215675354,
      "learning_rate": 1.8852313113324553e-05,
      "loss": 0.5833,
      "step": 1200
    },
    {
      "epoch": 2.42,
      "grad_norm": 1.700926661491394,
      "learning_rate": 1.8819623981158996e-05,
      "loss": 0.527,
      "step": 1210
    },
    {
      "epoch": 2.44,
      "grad_norm": 0.7639631032943726,
      "learning_rate": 1.878650499296323e-05,
      "loss": 0.5605,
      "step": 1220
    },
    {
      "epoch": 2.46,
      "grad_norm": 0.9155722856521606,
      "learning_rate": 1.8752957762910016e-05,
      "loss": 0.5528,
      "step": 1230
    },
    {
      "epoch": 2.48,
      "grad_norm": 1.4307914972305298,
      "learning_rate": 1.871898392604402e-05,
      "loss": 0.5239,
      "step": 1240
    },
    {
      "epoch": 2.5,
      "grad_norm": 0.9646509885787964,
      "learning_rate": 1.8684585138202122e-05,
      "loss": 0.5825,
      "step": 1250
    },
    {
      "epoch": 2.52,
      "grad_norm": 0.823187530040741,
      "learning_rate": 1.864976307593271e-05,
      "loss": 0.5816,
      "step": 1260
    },
    {
      "epoch": 2.54,
      "grad_norm": 0.9524659514427185,
      "learning_rate": 1.8614519436413968e-05,
      "loss": 0.595,
      "step": 1270
    },
    {
      "epoch": 2.56,
      "grad_norm": 3.476361036300659,
      "learning_rate": 1.8578855937371176e-05,
      "loss": 0.5828,
      "step": 1280
    },
    {
      "epoch": 2.58,
      "grad_norm": 0.8790251016616821,
      "learning_rate": 1.8542774316992953e-05,
      "loss": 0.5408,
      "step": 1290
    },
    {
      "epoch": 2.6,
      "grad_norm": 1.8205924034118652,
      "learning_rate": 1.850627633384658e-05,
      "loss": 0.5683,
      "step": 1300
    },
    {
      "epoch": 2.62,
      "grad_norm": 0.7987878918647766,
      "learning_rate": 1.8469363766792258e-05,
      "loss": 0.5734,
      "step": 1310
    },
    {
      "epoch": 2.64,
      "grad_norm": 1.1774803400039673,
      "learning_rate": 1.8432038414896432e-05,
      "loss": 0.5581,
      "step": 1320
    },
    {
      "epoch": 2.66,
      "grad_norm": 1.4109526872634888,
      "learning_rate": 1.8394302097344103e-05,
      "loss": 0.5781,
      "step": 1330
    },
    {
      "epoch": 2.68,
      "grad_norm": 2.104931116104126,
      "learning_rate": 1.8356156653350138e-05,
      "loss": 0.5468,
      "step": 1340
    },
    {
      "epoch": 2.7,
      "grad_norm": 1.7022193670272827,
      "learning_rate": 1.8317603942069665e-05,
      "loss": 0.543,
      "step": 1350
    },
    {
      "epoch": 2.7199999999999998,
      "grad_norm": 1.9831078052520752,
      "learning_rate": 1.8278645842507448e-05,
      "loss": 0.5416,
      "step": 1360
    },
    {
      "epoch": 2.74,
      "grad_norm": 1.4346486330032349,
      "learning_rate": 1.8239284253426294e-05,
      "loss": 0.5692,
      "step": 1370
    },
    {
      "epoch": 2.76,
      "grad_norm": 3.786205768585205,
      "learning_rate": 1.8199521093254524e-05,
      "loss": 0.5372,
      "step": 1380
    },
    {
      "epoch": 2.7800000000000002,
      "grad_norm": 1.2855556011199951,
      "learning_rate": 1.815935829999247e-05,
      "loss": 0.5205,
      "step": 1390
    },
    {
      "epoch": 2.8,
      "grad_norm": 1.4858529567718506,
      "learning_rate": 1.811879783111801e-05,
      "loss": 0.5159,
      "step": 1400
    },
    {
      "epoch": 2.82,
      "grad_norm": 1.3300436735153198,
      "learning_rate": 1.8077841663491174e-05,
      "loss": 0.5405,
      "step": 1410
    },
    {
      "epoch": 2.84,
      "grad_norm": 1.2626831531524658,
      "learning_rate": 1.80364917932578e-05,
      "loss": 0.5769,
      "step": 1420
    },
    {
      "epoch": 2.86,
      "grad_norm": 1.4615288972854614,
      "learning_rate": 1.799475023575222e-05,
      "loss": 0.5724,
      "step": 1430
    },
    {
      "epoch": 2.88,
      "grad_norm": 1.2775332927703857,
      "learning_rate": 1.795261902539906e-05,
      "loss": 0.5603,
      "step": 1440
    },
    {
      "epoch": 2.9,
      "grad_norm": 1.9211925268173218,
      "learning_rate": 1.791010021561407e-05,
      "loss": 0.5609,
      "step": 1450
    },
    {
      "epoch": 2.92,
      "grad_norm": 3.850076675415039,
      "learning_rate": 1.7867195878704062e-05,
      "loss": 0.585,
      "step": 1460
    },
    {
      "epoch": 2.94,
      "grad_norm": 1.0455843210220337,
      "learning_rate": 1.7823908105765883e-05,
      "loss": 0.5818,
      "step": 1470
    },
    {
      "epoch": 2.96,
      "grad_norm": 2.406511068344116,
      "learning_rate": 1.7780239006584515e-05,
      "loss": 0.5453,
      "step": 1480
    },
    {
      "epoch": 2.98,
      "grad_norm": 1.212849736213684,
      "learning_rate": 1.773619070953025e-05,
      "loss": 0.5526,
      "step": 1490
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.608502984046936,
      "learning_rate": 1.769176536145494e-05,
      "loss": 0.5664,
      "step": 1500
    },
    {
      "epoch": 3.02,
      "grad_norm": 2.083935499191284,
      "learning_rate": 1.7646965127587373e-05,
      "loss": 0.4993,
      "step": 1510
    },
    {
      "epoch": 3.04,
      "grad_norm": 3.191681146621704,
      "learning_rate": 1.760179219142774e-05,
      "loss": 0.5302,
      "step": 1520
    },
    {
      "epoch": 3.06,
      "grad_norm": 2.241614580154419,
      "learning_rate": 1.7556248754641237e-05,
      "loss": 0.4995,
      "step": 1530
    },
    {
      "epoch": 3.08,
      "grad_norm": 2.2549211978912354,
      "learning_rate": 1.7510337036950703e-05,
      "loss": 0.4902,
      "step": 1540
    },
    {
      "epoch": 3.1,
      "grad_norm": 2.623849630355835,
      "learning_rate": 1.7464059276028497e-05,
      "loss": 0.5571,
      "step": 1550
    },
    {
      "epoch": 3.12,
      "grad_norm": 1.099225401878357,
      "learning_rate": 1.7417417727387392e-05,
      "loss": 0.5441,
      "step": 1560
    },
    {
      "epoch": 3.14,
      "grad_norm": 1.0318654775619507,
      "learning_rate": 1.7370414664270675e-05,
      "loss": 0.5498,
      "step": 1570
    },
    {
      "epoch": 3.16,
      "grad_norm": 1.4295272827148438,
      "learning_rate": 1.732305237754132e-05,
      "loss": 0.4799,
      "step": 1580
    },
    {
      "epoch": 3.18,
      "grad_norm": 3.886106014251709,
      "learning_rate": 1.727533317557037e-05,
      "loss": 0.5285,
      "step": 1590
    },
    {
      "epoch": 3.2,
      "grad_norm": 0.9527933597564697,
      "learning_rate": 1.7227259384124408e-05,
      "loss": 0.5328,
      "step": 1600
    },
    {
      "epoch": 3.22,
      "grad_norm": 1.0082141160964966,
      "learning_rate": 1.7178833346252208e-05,
      "loss": 0.5333,
      "step": 1610
    },
    {
      "epoch": 3.24,
      "grad_norm": 0.9779771566390991,
      "learning_rate": 1.713005742217053e-05,
      "loss": 0.5163,
      "step": 1620
    },
    {
      "epoch": 3.26,
      "grad_norm": 2.230241060256958,
      "learning_rate": 1.7080933989149112e-05,
      "loss": 0.5173,
      "step": 1630
    },
    {
      "epoch": 3.2800000000000002,
      "grad_norm": 4.50337028503418,
      "learning_rate": 1.7031465441394766e-05,
      "loss": 0.5187,
      "step": 1640
    },
    {
      "epoch": 3.3,
      "grad_norm": 1.1370171308517456,
      "learning_rate": 1.698165418993473e-05,
      "loss": 0.5611,
      "step": 1650
    },
    {
      "epoch": 3.32,
      "grad_norm": 0.9167156219482422,
      "learning_rate": 1.6931502662499116e-05,
      "loss": 0.5381,
      "step": 1660
    },
    {
      "epoch": 3.34,
      "grad_norm": 2.5473690032958984,
      "learning_rate": 1.688101330340263e-05,
      "loss": 0.5089,
      "step": 1670
    },
    {
      "epoch": 3.36,
      "grad_norm": 0.9971883893013,
      "learning_rate": 1.683018857342539e-05,
      "loss": 0.5538,
      "step": 1680
    },
    {
      "epoch": 3.38,
      "grad_norm": 1.7401093244552612,
      "learning_rate": 1.6779030949693044e-05,
      "loss": 0.5216,
      "step": 1690
    },
    {
      "epoch": 3.4,
      "grad_norm": 6.5078935623168945,
      "learning_rate": 1.6727542925556e-05,
      "loss": 0.5356,
      "step": 1700
    },
    {
      "epoch": 3.42,
      "grad_norm": 0.9460225105285645,
      "learning_rate": 1.667572701046791e-05,
      "loss": 0.497,
      "step": 1710
    },
    {
      "epoch": 3.44,
      "grad_norm": 1.153794765472412,
      "learning_rate": 1.662358572986337e-05,
      "loss": 0.4934,
      "step": 1720
    },
    {
      "epoch": 3.46,
      "grad_norm": 1.1257123947143555,
      "learning_rate": 1.6571121625034847e-05,
      "loss": 0.5327,
      "step": 1730
    },
    {
      "epoch": 3.48,
      "grad_norm": 2.454582691192627,
      "learning_rate": 1.651833725300879e-05,
      "loss": 0.4995,
      "step": 1740
    },
    {
      "epoch": 3.5,
      "grad_norm": 1.7266925573349,
      "learning_rate": 1.6465235186421024e-05,
      "loss": 0.4945,
      "step": 1750
    },
    {
      "epoch": 3.52,
      "grad_norm": 2.5186126232147217,
      "learning_rate": 1.6411818013391357e-05,
      "loss": 0.4969,
      "step": 1760
    },
    {
      "epoch": 3.54,
      "grad_norm": 2.2978458404541016,
      "learning_rate": 1.6358088337397444e-05,
      "loss": 0.5133,
      "step": 1770
    },
    {
      "epoch": 3.56,
      "grad_norm": 9.083796501159668,
      "learning_rate": 1.630404877714789e-05,
      "loss": 0.4598,
      "step": 1780
    },
    {
      "epoch": 3.58,
      "grad_norm": 1.687730312347412,
      "learning_rate": 1.6249701966454626e-05,
      "loss": 0.5721,
      "step": 1790
    },
    {
      "epoch": 3.6,
      "grad_norm": 0.949276328086853,
      "learning_rate": 1.619505055410453e-05,
      "loss": 0.5549,
      "step": 1800
    },
    {
      "epoch": 3.62,
      "grad_norm": 1.7653878927230835,
      "learning_rate": 1.614009720373034e-05,
      "loss": 0.5192,
      "step": 1810
    },
    {
      "epoch": 3.64,
      "grad_norm": 0.6442993879318237,
      "learning_rate": 1.608484459368082e-05,
      "loss": 0.4927,
      "step": 1820
    },
    {
      "epoch": 3.66,
      "grad_norm": 2.4791717529296875,
      "learning_rate": 1.602929541689025e-05,
      "loss": 0.5319,
      "step": 1830
    },
    {
      "epoch": 3.68,
      "grad_norm": 1.8095901012420654,
      "learning_rate": 1.5973452380747125e-05,
      "loss": 0.5025,
      "step": 1840
    },
    {
      "epoch": 3.7,
      "grad_norm": 1.8692318201065063,
      "learning_rate": 1.591731820696224e-05,
      "loss": 0.497,
      "step": 1850
    },
    {
      "epoch": 3.7199999999999998,
      "grad_norm": 1.3001285791397095,
      "learning_rate": 1.5860895631436044e-05,
      "loss": 0.556,
      "step": 1860
    },
    {
      "epoch": 3.74,
      "grad_norm": 1.0697414875030518,
      "learning_rate": 1.580418740412526e-05,
      "loss": 0.506,
      "step": 1870
    },
    {
      "epoch": 3.76,
      "grad_norm": 3.5158541202545166,
      "learning_rate": 1.5747196288908887e-05,
      "loss": 0.5154,
      "step": 1880
    },
    {
      "epoch": 3.7800000000000002,
      "grad_norm": 1.699308156967163,
      "learning_rate": 1.5689925063453483e-05,
      "loss": 0.5887,
      "step": 1890
    },
    {
      "epoch": 3.8,
      "grad_norm": 0.825036883354187,
      "learning_rate": 1.563237651907777e-05,
      "loss": 0.508,
      "step": 1900
    },
    {
      "epoch": 3.82,
      "grad_norm": 1.7058846950531006,
      "learning_rate": 1.5574553460616608e-05,
      "loss": 0.4954,
      "step": 1910
    },
    {
      "epoch": 3.84,
      "grad_norm": 1.7128701210021973,
      "learning_rate": 1.5516458706284306e-05,
      "loss": 0.5628,
      "step": 1920
    },
    {
      "epoch": 3.86,
      "grad_norm": 0.8009471297264099,
      "learning_rate": 1.5458095087537216e-05,
      "loss": 0.4494,
      "step": 1930
    },
    {
      "epoch": 3.88,
      "grad_norm": 1.8152306079864502,
      "learning_rate": 1.5399465448935788e-05,
      "loss": 0.522,
      "step": 1940
    },
    {
      "epoch": 3.9,
      "grad_norm": 2.59840989112854,
      "learning_rate": 1.5340572648005887e-05,
      "loss": 0.5225,
      "step": 1950
    },
    {
      "epoch": 3.92,
      "grad_norm": 3.8816378116607666,
      "learning_rate": 1.5281419555099547e-05,
      "loss": 0.5092,
      "step": 1960
    },
    {
      "epoch": 3.94,
      "grad_norm": 1.4815788269042969,
      "learning_rate": 1.5222009053255061e-05,
      "loss": 0.5167,
      "step": 1970
    },
    {
      "epoch": 3.96,
      "grad_norm": 1.5924495458602905,
      "learning_rate": 1.5162344038056476e-05,
      "loss": 0.5198,
      "step": 1980
    },
    {
      "epoch": 3.98,
      "grad_norm": 1.44657301902771,
      "learning_rate": 1.510242741749246e-05,
      "loss": 0.5723,
      "step": 1990
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.8102152347564697,
      "learning_rate": 1.5042262111814566e-05,
      "loss": 0.4707,
      "step": 2000
    },
    {
      "epoch": 4.02,
      "grad_norm": 1.1032963991165161,
      "learning_rate": 1.498185105339491e-05,
      "loss": 0.523,
      "step": 2010
    },
    {
      "epoch": 4.04,
      "grad_norm": 1.0523751974105835,
      "learning_rate": 1.4921197186583256e-05,
      "loss": 0.4433,
      "step": 2020
    },
    {
      "epoch": 4.06,
      "grad_norm": 3.989997148513794,
      "learning_rate": 1.4860303467563504e-05,
      "loss": 0.4861,
      "step": 2030
    },
    {
      "epoch": 4.08,
      "grad_norm": 2.7621233463287354,
      "learning_rate": 1.4799172864209607e-05,
      "loss": 0.4621,
      "step": 2040
    },
    {
      "epoch": 4.1,
      "grad_norm": 0.960752010345459,
      "learning_rate": 1.4737808355940932e-05,
      "loss": 0.5454,
      "step": 2050
    },
    {
      "epoch": 4.12,
      "grad_norm": 1.279720664024353,
      "learning_rate": 1.467621293357704e-05,
      "loss": 0.4984,
      "step": 2060
    },
    {
      "epoch": 4.14,
      "grad_norm": 2.1226606369018555,
      "learning_rate": 1.4614389599191917e-05,
      "loss": 0.5375,
      "step": 2070
    },
    {
      "epoch": 4.16,
      "grad_norm": 1.6314127445220947,
      "learning_rate": 1.455234136596766e-05,
      "loss": 0.5191,
      "step": 2080
    },
    {
      "epoch": 4.18,
      "grad_norm": 2.72589111328125,
      "learning_rate": 1.4490071258047625e-05,
      "loss": 0.4872,
      "step": 2090
    },
    {
      "epoch": 4.2,
      "grad_norm": 4.3185648918151855,
      "learning_rate": 1.442758231038902e-05,
      "loss": 0.4883,
      "step": 2100
    },
    {
      "epoch": 4.22,
      "grad_norm": 0.8917685747146606,
      "learning_rate": 1.436487756861499e-05,
      "loss": 0.4564,
      "step": 2110
    },
    {
      "epoch": 4.24,
      "grad_norm": 1.1269738674163818,
      "learning_rate": 1.4301960088866187e-05,
      "loss": 0.4311,
      "step": 2120
    },
    {
      "epoch": 4.26,
      "grad_norm": 4.341186046600342,
      "learning_rate": 1.4238832937651816e-05,
      "loss": 0.4663,
      "step": 2130
    },
    {
      "epoch": 4.28,
      "grad_norm": 0.9497487545013428,
      "learning_rate": 1.4175499191700169e-05,
      "loss": 0.5048,
      "step": 2140
    },
    {
      "epoch": 4.3,
      "grad_norm": 1.7001278400421143,
      "learning_rate": 1.4111961937808665e-05,
      "loss": 0.493,
      "step": 2150
    },
    {
      "epoch": 4.32,
      "grad_norm": 1.7037960290908813,
      "learning_rate": 1.4048224272693426e-05,
      "loss": 0.4712,
      "step": 2160
    },
    {
      "epoch": 4.34,
      "grad_norm": 0.9490267634391785,
      "learning_rate": 1.3984289302838327e-05,
      "loss": 0.4641,
      "step": 2170
    },
    {
      "epoch": 4.36,
      "grad_norm": 2.2001664638519287,
      "learning_rate": 1.3920160144343604e-05,
      "loss": 0.4929,
      "step": 2180
    },
    {
      "epoch": 4.38,
      "grad_norm": 1.0687891244888306,
      "learning_rate": 1.3855839922773968e-05,
      "loss": 0.5269,
      "step": 2190
    },
    {
      "epoch": 4.4,
      "grad_norm": 1.1579177379608154,
      "learning_rate": 1.3791331773006272e-05,
      "loss": 0.4857,
      "step": 2200
    },
    {
      "epoch": 4.42,
      "grad_norm": 0.9180253744125366,
      "learning_rate": 1.3726638839076732e-05,
      "loss": 0.5613,
      "step": 2210
    },
    {
      "epoch": 4.44,
      "grad_norm": 1.170154333114624,
      "learning_rate": 1.3661764274027678e-05,
      "loss": 0.4884,
      "step": 2220
    },
    {
      "epoch": 4.46,
      "grad_norm": 1.8389673233032227,
      "learning_rate": 1.3596711239753889e-05,
      "loss": 0.4849,
      "step": 2230
    },
    {
      "epoch": 4.48,
      "grad_norm": 1.43435537815094,
      "learning_rate": 1.3531482906848474e-05,
      "loss": 0.4752,
      "step": 2240
    },
    {
      "epoch": 4.5,
      "grad_norm": 1.2561147212982178,
      "learning_rate": 1.3466082454448364e-05,
      "loss": 0.4804,
      "step": 2250
    },
    {
      "epoch": 4.52,
      "grad_norm": 1.3098878860473633,
      "learning_rate": 1.340051307007933e-05,
      "loss": 0.4719,
      "step": 2260
    },
    {
      "epoch": 4.54,
      "grad_norm": 7.966071128845215,
      "learning_rate": 1.3334777949500673e-05,
      "loss": 0.4599,
      "step": 2270
    },
    {
      "epoch": 4.5600000000000005,
      "grad_norm": 2.28067946434021,
      "learning_rate": 1.3268880296549424e-05,
      "loss": 0.4712,
      "step": 2280
    },
    {
      "epoch": 4.58,
      "grad_norm": 1.1881040334701538,
      "learning_rate": 1.3202823322984228e-05,
      "loss": 0.4772,
      "step": 2290
    },
    {
      "epoch": 4.6,
      "grad_norm": 1.482059359550476,
      "learning_rate": 1.3136610248328779e-05,
      "loss": 0.453,
      "step": 2300
    },
    {
      "epoch": 4.62,
      "grad_norm": 2.1320247650146484,
      "learning_rate": 1.307024429971492e-05,
      "loss": 0.4657,
      "step": 2310
    },
    {
      "epoch": 4.64,
      "grad_norm": 1.2124568223953247,
      "learning_rate": 1.3003728711725364e-05,
      "loss": 0.4791,
      "step": 2320
    },
    {
      "epoch": 4.66,
      "grad_norm": 0.7249093055725098,
      "learning_rate": 1.2937066726236029e-05,
      "loss": 0.5586,
      "step": 2330
    },
    {
      "epoch": 4.68,
      "grad_norm": 1.5038686990737915,
      "learning_rate": 1.2870261592258038e-05,
      "loss": 0.4603,
      "step": 2340
    },
    {
      "epoch": 4.7,
      "grad_norm": 1.320827603340149,
      "learning_rate": 1.2803316565779378e-05,
      "loss": 0.4618,
      "step": 2350
    },
    {
      "epoch": 4.72,
      "grad_norm": 1.4801580905914307,
      "learning_rate": 1.2736234909606186e-05,
      "loss": 0.4643,
      "step": 2360
    },
    {
      "epoch": 4.74,
      "grad_norm": 2.649113655090332,
      "learning_rate": 1.2669019893203758e-05,
      "loss": 0.5017,
      "step": 2370
    },
    {
      "epoch": 4.76,
      "grad_norm": 1.3265902996063232,
      "learning_rate": 1.2601674792537157e-05,
      "loss": 0.451,
      "step": 2380
    },
    {
      "epoch": 4.78,
      "grad_norm": 1.6874167919158936,
      "learning_rate": 1.2534202889911584e-05,
      "loss": 0.4763,
      "step": 2390
    },
    {
      "epoch": 4.8,
      "grad_norm": 0.7948456406593323,
      "learning_rate": 1.2466607473812386e-05,
      "loss": 0.4984,
      "step": 2400
    },
    {
      "epoch": 4.82,
      "grad_norm": 1.3831831216812134,
      "learning_rate": 1.2398891838744777e-05,
      "loss": 0.4594,
      "step": 2410
    },
    {
      "epoch": 4.84,
      "grad_norm": 2.660630941390991,
      "learning_rate": 1.233105928507328e-05,
      "loss": 0.476,
      "step": 2420
    },
    {
      "epoch": 4.86,
      "grad_norm": 1.7315685749053955,
      "learning_rate": 1.226311311886086e-05,
      "loss": 0.4599,
      "step": 2430
    },
    {
      "epoch": 4.88,
      "grad_norm": 1.0656920671463013,
      "learning_rate": 1.2195056651707806e-05,
      "loss": 0.4786,
      "step": 2440
    },
    {
      "epoch": 4.9,
      "grad_norm": 1.317185878753662,
      "learning_rate": 1.2126893200590309e-05,
      "loss": 0.539,
      "step": 2450
    },
    {
      "epoch": 4.92,
      "grad_norm": 2.1588151454925537,
      "learning_rate": 1.2058626087698814e-05,
      "loss": 0.442,
      "step": 2460
    },
    {
      "epoch": 4.9399999999999995,
      "grad_norm": 1.3337817192077637,
      "learning_rate": 1.1990258640276094e-05,
      "loss": 0.4829,
      "step": 2470
    },
    {
      "epoch": 4.96,
      "grad_norm": 1.3558602333068848,
      "learning_rate": 1.1921794190455082e-05,
      "loss": 0.5055,
      "step": 2480
    },
    {
      "epoch": 4.98,
      "grad_norm": 1.18630850315094,
      "learning_rate": 1.1853236075096474e-05,
      "loss": 0.4857,
      "step": 2490
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.4073606729507446,
      "learning_rate": 1.1784587635626095e-05,
      "loss": 0.4962,
      "step": 2500
    },
    {
      "epoch": 5.02,
      "grad_norm": 2.1152431964874268,
      "learning_rate": 1.171585221787203e-05,
      "loss": 0.4185,
      "step": 2510
    },
    {
      "epoch": 5.04,
      "grad_norm": 1.4434751272201538,
      "learning_rate": 1.1647033171901573e-05,
      "loss": 0.4545,
      "step": 2520
    },
    {
      "epoch": 5.06,
      "grad_norm": 2.575100898742676,
      "learning_rate": 1.157813385185794e-05,
      "loss": 0.4162,
      "step": 2530
    },
    {
      "epoch": 5.08,
      "grad_norm": 1.808670163154602,
      "learning_rate": 1.1509157615796775e-05,
      "loss": 0.425,
      "step": 2540
    },
    {
      "epoch": 5.1,
      "grad_norm": 2.0756947994232178,
      "learning_rate": 1.1440107825522522e-05,
      "loss": 0.4514,
      "step": 2550
    },
    {
      "epoch": 5.12,
      "grad_norm": 1.3622123003005981,
      "learning_rate": 1.1370987846424547e-05,
      "loss": 0.4687,
      "step": 2560
    },
    {
      "epoch": 5.14,
      "grad_norm": 1.938477873802185,
      "learning_rate": 1.1301801047313106e-05,
      "loss": 0.4892,
      "step": 2570
    },
    {
      "epoch": 5.16,
      "grad_norm": 1.3552794456481934,
      "learning_rate": 1.1232550800255188e-05,
      "loss": 0.4675,
      "step": 2580
    },
    {
      "epoch": 5.18,
      "grad_norm": 2.3927013874053955,
      "learning_rate": 1.1163240480410136e-05,
      "loss": 0.4336,
      "step": 2590
    },
    {
      "epoch": 5.2,
      "grad_norm": 1.3408219814300537,
      "learning_rate": 1.1093873465865156e-05,
      "loss": 0.4358,
      "step": 2600
    },
    {
      "epoch": 5.22,
      "grad_norm": 2.9869275093078613,
      "learning_rate": 1.1024453137470677e-05,
      "loss": 0.4709,
      "step": 2610
    },
    {
      "epoch": 5.24,
      "grad_norm": 2.7815663814544678,
      "learning_rate": 1.0954982878675564e-05,
      "loss": 0.4349,
      "step": 2620
    },
    {
      "epoch": 5.26,
      "grad_norm": 1.9527360200881958,
      "learning_rate": 1.0885466075362224e-05,
      "loss": 0.4581,
      "step": 2630
    },
    {
      "epoch": 5.28,
      "grad_norm": 1.804969072341919,
      "learning_rate": 1.0815906115681579e-05,
      "loss": 0.4482,
      "step": 2640
    },
    {
      "epoch": 5.3,
      "grad_norm": 1.6230815649032593,
      "learning_rate": 1.0746306389887924e-05,
      "loss": 0.4771,
      "step": 2650
    },
    {
      "epoch": 5.32,
      "grad_norm": 2.6288340091705322,
      "learning_rate": 1.067667029017371e-05,
      "loss": 0.4893,
      "step": 2660
    },
    {
      "epoch": 5.34,
      "grad_norm": 0.7628895044326782,
      "learning_rate": 1.060700121050419e-05,
      "loss": 0.4823,
      "step": 2670
    },
    {
      "epoch": 5.36,
      "grad_norm": 1.395524501800537,
      "learning_rate": 1.0537302546452022e-05,
      "loss": 0.45,
      "step": 2680
    },
    {
      "epoch": 5.38,
      "grad_norm": 1.0264369249343872,
      "learning_rate": 1.0467577695031763e-05,
      "loss": 0.4817,
      "step": 2690
    },
    {
      "epoch": 5.4,
      "grad_norm": 1.3651304244995117,
      "learning_rate": 1.03978300545343e-05,
      "loss": 0.4472,
      "step": 2700
    },
    {
      "epoch": 5.42,
      "grad_norm": 1.520727276802063,
      "learning_rate": 1.0328063024361232e-05,
      "loss": 0.4351,
      "step": 2710
    },
    {
      "epoch": 5.44,
      "grad_norm": 2.283327102661133,
      "learning_rate": 1.0258280004859189e-05,
      "loss": 0.4052,
      "step": 2720
    },
    {
      "epoch": 5.46,
      "grad_norm": 5.243598937988281,
      "learning_rate": 1.0188484397154083e-05,
      "loss": 0.51,
      "step": 2730
    },
    {
      "epoch": 5.48,
      "grad_norm": 2.3326563835144043,
      "learning_rate": 1.0118679602985373e-05,
      "loss": 0.4678,
      "step": 2740
    },
    {
      "epoch": 5.5,
      "grad_norm": 1.8756747245788574,
      "learning_rate": 1.0048869024540247e-05,
      "loss": 0.4802,
      "step": 2750
    },
    {
      "epoch": 5.52,
      "grad_norm": 2.212642192840576,
      "learning_rate": 9.979056064287807e-06,
      "loss": 0.4416,
      "step": 2760
    },
    {
      "epoch": 5.54,
      "grad_norm": 1.893557071685791,
      "learning_rate": 9.909244124813246e-06,
      "loss": 0.4613,
      "step": 2770
    },
    {
      "epoch": 5.5600000000000005,
      "grad_norm": 3.211782217025757,
      "learning_rate": 9.839436608652007e-06,
      "loss": 0.4163,
      "step": 2780
    },
    {
      "epoch": 5.58,
      "grad_norm": 1.4164925813674927,
      "learning_rate": 9.76963691812394e-06,
      "loss": 0.4753,
      "step": 2790
    },
    {
      "epoch": 5.6,
      "grad_norm": 1.139273762702942,
      "learning_rate": 9.699848455167489e-06,
      "loss": 0.4725,
      "step": 2800
    },
    {
      "epoch": 5.62,
      "grad_norm": 1.5774643421173096,
      "learning_rate": 9.630074621173882e-06,
      "loss": 0.4521,
      "step": 2810
    },
    {
      "epoch": 5.64,
      "grad_norm": 2.0061256885528564,
      "learning_rate": 9.560318816821354e-06,
      "loss": 0.3838,
      "step": 2820
    },
    {
      "epoch": 5.66,
      "grad_norm": 3.7671396732330322,
      "learning_rate": 9.490584441909392e-06,
      "loss": 0.4603,
      "step": 2830
    },
    {
      "epoch": 5.68,
      "grad_norm": 1.6197257041931152,
      "learning_rate": 9.420874895193056e-06,
      "loss": 0.4538,
      "step": 2840
    },
    {
      "epoch": 5.7,
      "grad_norm": 3.386794328689575,
      "learning_rate": 9.351193574217305e-06,
      "loss": 0.4527,
      "step": 2850
    },
    {
      "epoch": 5.72,
      "grad_norm": 1.4062769412994385,
      "learning_rate": 9.281543875151419e-06,
      "loss": 0.4915,
      "step": 2860
    },
    {
      "epoch": 5.74,
      "grad_norm": 2.367417573928833,
      "learning_rate": 9.211929192623466e-06,
      "loss": 0.4338,
      "step": 2870
    },
    {
      "epoch": 5.76,
      "grad_norm": 1.7326956987380981,
      "learning_rate": 9.142352919554862e-06,
      "loss": 0.4573,
      "step": 2880
    },
    {
      "epoch": 5.78,
      "grad_norm": 2.8575878143310547,
      "learning_rate": 9.072818446995e-06,
      "loss": 0.4494,
      "step": 2890
    },
    {
      "epoch": 5.8,
      "grad_norm": 1.1295793056488037,
      "learning_rate": 9.003329163955973e-06,
      "loss": 0.5061,
      "step": 2900
    },
    {
      "epoch": 5.82,
      "grad_norm": 1.31191885471344,
      "learning_rate": 8.933888457247402e-06,
      "loss": 0.4537,
      "step": 2910
    },
    {
      "epoch": 5.84,
      "grad_norm": 1.8330936431884766,
      "learning_rate": 8.864499711311362e-06,
      "loss": 0.4764,
      "step": 2920
    },
    {
      "epoch": 5.86,
      "grad_norm": 1.8839877843856812,
      "learning_rate": 8.79516630805745e-06,
      "loss": 0.4563,
      "step": 2930
    },
    {
      "epoch": 5.88,
      "grad_norm": 2.6970791816711426,
      "learning_rate": 8.725891626697912e-06,
      "loss": 0.4887,
      "step": 2940
    },
    {
      "epoch": 5.9,
      "grad_norm": 3.178072214126587,
      "learning_rate": 8.656679043582986e-06,
      "loss": 0.446,
      "step": 2950
    },
    {
      "epoch": 5.92,
      "grad_norm": 2.423067569732666,
      "learning_rate": 8.587531932036334e-06,
      "loss": 0.4533,
      "step": 2960
    },
    {
      "epoch": 5.9399999999999995,
      "grad_norm": 3.043440580368042,
      "learning_rate": 8.518453662190622e-06,
      "loss": 0.4451,
      "step": 2970
    },
    {
      "epoch": 5.96,
      "grad_norm": 2.4324257373809814,
      "learning_rate": 8.449447600823262e-06,
      "loss": 0.393,
      "step": 2980
    },
    {
      "epoch": 5.98,
      "grad_norm": 7.399738311767578,
      "learning_rate": 8.380517111192336e-06,
      "loss": 0.4406,
      "step": 2990
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.8923618197441101,
      "learning_rate": 8.311665552872662e-06,
      "loss": 0.474,
      "step": 3000
    },
    {
      "epoch": 6.02,
      "grad_norm": 1.7500466108322144,
      "learning_rate": 8.242896281592057e-06,
      "loss": 0.3953,
      "step": 3010
    },
    {
      "epoch": 6.04,
      "grad_norm": 3.8324530124664307,
      "learning_rate": 8.174212649067781e-06,
      "loss": 0.4117,
      "step": 3020
    },
    {
      "epoch": 6.06,
      "grad_norm": 2.275822639465332,
      "learning_rate": 8.10561800284319e-06,
      "loss": 0.3988,
      "step": 3030
    },
    {
      "epoch": 6.08,
      "grad_norm": 1.2562943696975708,
      "learning_rate": 8.037115686124564e-06,
      "loss": 0.418,
      "step": 3040
    },
    {
      "epoch": 6.1,
      "grad_norm": 1.3214370012283325,
      "learning_rate": 7.96870903761818e-06,
      "loss": 0.4084,
      "step": 3050
    },
    {
      "epoch": 6.12,
      "grad_norm": 2.595797061920166,
      "learning_rate": 7.900401391367576e-06,
      "loss": 0.3739,
      "step": 3060
    },
    {
      "epoch": 6.14,
      "grad_norm": 2.055779457092285,
      "learning_rate": 7.832196076591067e-06,
      "loss": 0.3763,
      "step": 3070
    },
    {
      "epoch": 6.16,
      "grad_norm": 2.5182206630706787,
      "learning_rate": 7.76409641751947e-06,
      "loss": 0.4522,
      "step": 3080
    },
    {
      "epoch": 6.18,
      "grad_norm": 6.86693000793457,
      "learning_rate": 7.696105733234099e-06,
      "loss": 0.4661,
      "step": 3090
    },
    {
      "epoch": 6.2,
      "grad_norm": 0.7651225924491882,
      "learning_rate": 7.628227337504972e-06,
      "loss": 0.4519,
      "step": 3100
    },
    {
      "epoch": 6.22,
      "grad_norm": 1.531447172164917,
      "learning_rate": 7.560464538629345e-06,
      "loss": 0.4073,
      "step": 3110
    },
    {
      "epoch": 6.24,
      "grad_norm": 2.4921135902404785,
      "learning_rate": 7.492820639270435e-06,
      "loss": 0.4458,
      "step": 3120
    },
    {
      "epoch": 6.26,
      "grad_norm": 1.5849100351333618,
      "learning_rate": 7.4252989362964635e-06,
      "loss": 0.3703,
      "step": 3130
    },
    {
      "epoch": 6.28,
      "grad_norm": 1.8190685510635376,
      "learning_rate": 7.357902720619976e-06,
      "loss": 0.4393,
      "step": 3140
    },
    {
      "epoch": 6.3,
      "grad_norm": 1.195379376411438,
      "learning_rate": 7.290635277037442e-06,
      "loss": 0.437,
      "step": 3150
    },
    {
      "epoch": 6.32,
      "grad_norm": 1.6389209032058716,
      "learning_rate": 7.22349988406916e-06,
      "loss": 0.3979,
      "step": 3160
    },
    {
      "epoch": 6.34,
      "grad_norm": 1.7568351030349731,
      "learning_rate": 7.156499813799477e-06,
      "loss": 0.4078,
      "step": 3170
    },
    {
      "epoch": 6.36,
      "grad_norm": 2.5609893798828125,
      "learning_rate": 7.0896383317172845e-06,
      "loss": 0.4182,
      "step": 3180
    },
    {
      "epoch": 6.38,
      "grad_norm": 2.070969343185425,
      "learning_rate": 7.022918696556896e-06,
      "loss": 0.4239,
      "step": 3190
    },
    {
      "epoch": 6.4,
      "grad_norm": 3.3474345207214355,
      "learning_rate": 6.956344160139201e-06,
      "loss": 0.4369,
      "step": 3200
    },
    {
      "epoch": 6.42,
      "grad_norm": 1.3559445142745972,
      "learning_rate": 6.889917967213184e-06,
      "loss": 0.4469,
      "step": 3210
    },
    {
      "epoch": 6.44,
      "grad_norm": 1.4630825519561768,
      "learning_rate": 6.823643355297774e-06,
      "loss": 0.4312,
      "step": 3220
    },
    {
      "epoch": 6.46,
      "grad_norm": 2.0589451789855957,
      "learning_rate": 6.757523554524056e-06,
      "loss": 0.4465,
      "step": 3230
    },
    {
      "epoch": 6.48,
      "grad_norm": 2.448317766189575,
      "learning_rate": 6.69156178747784e-06,
      "loss": 0.4201,
      "step": 3240
    },
    {
      "epoch": 6.5,
      "grad_norm": 1.911927580833435,
      "learning_rate": 6.62576126904259e-06,
      "loss": 0.3882,
      "step": 3250
    },
    {
      "epoch": 6.52,
      "grad_norm": 3.176950216293335,
      "learning_rate": 6.560125206242746e-06,
      "loss": 0.4448,
      "step": 3260
    },
    {
      "epoch": 6.54,
      "grad_norm": 1.4145492315292358,
      "learning_rate": 6.494656798087412e-06,
      "loss": 0.3915,
      "step": 3270
    },
    {
      "epoch": 6.5600000000000005,
      "grad_norm": 4.982487201690674,
      "learning_rate": 6.4293592354144365e-06,
      "loss": 0.3769,
      "step": 3280
    },
    {
      "epoch": 6.58,
      "grad_norm": 3.9301717281341553,
      "learning_rate": 6.364235700734903e-06,
      "loss": 0.4503,
      "step": 3290
    },
    {
      "epoch": 6.6,
      "grad_norm": 3.643587112426758,
      "learning_rate": 6.299289368078016e-06,
      "loss": 0.4398,
      "step": 3300
    },
    {
      "epoch": 6.62,
      "grad_norm": 2.1195595264434814,
      "learning_rate": 6.234523402836408e-06,
      "loss": 0.4199,
      "step": 3310
    },
    {
      "epoch": 6.64,
      "grad_norm": 1.3784760236740112,
      "learning_rate": 6.169940961611853e-06,
      "loss": 0.4574,
      "step": 3320
    },
    {
      "epoch": 6.66,
      "grad_norm": 4.373683452606201,
      "learning_rate": 6.1055451920614165e-06,
      "loss": 0.4252,
      "step": 3330
    },
    {
      "epoch": 6.68,
      "grad_norm": 1.9143246412277222,
      "learning_rate": 6.0413392327440635e-06,
      "loss": 0.4069,
      "step": 3340
    },
    {
      "epoch": 6.7,
      "grad_norm": 1.9608592987060547,
      "learning_rate": 5.977326212967671e-06,
      "loss": 0.4173,
      "step": 3350
    },
    {
      "epoch": 6.72,
      "grad_norm": 2.7381646633148193,
      "learning_rate": 5.913509252636511e-06,
      "loss": 0.3737,
      "step": 3360
    },
    {
      "epoch": 6.74,
      "grad_norm": 1.9294111728668213,
      "learning_rate": 5.849891462099199e-06,
      "loss": 0.437,
      "step": 3370
    },
    {
      "epoch": 6.76,
      "grad_norm": 0.8937060236930847,
      "learning_rate": 5.786475941997094e-06,
      "loss": 0.4457,
      "step": 3380
    },
    {
      "epoch": 6.78,
      "grad_norm": 1.2962634563446045,
      "learning_rate": 5.723265783113181e-06,
      "loss": 0.3989,
      "step": 3390
    },
    {
      "epoch": 6.8,
      "grad_norm": 1.0391006469726562,
      "learning_rate": 5.660264066221426e-06,
      "loss": 0.4314,
      "step": 3400
    },
    {
      "epoch": 6.82,
      "grad_norm": 7.157230377197266,
      "learning_rate": 5.59747386193663e-06,
      "loss": 0.3989,
      "step": 3410
    },
    {
      "epoch": 6.84,
      "grad_norm": 2.1349549293518066,
      "learning_rate": 5.534898230564765e-06,
      "loss": 0.3792,
      "step": 3420
    },
    {
      "epoch": 6.86,
      "grad_norm": 1.362468957901001,
      "learning_rate": 5.472540221953824e-06,
      "loss": 0.4115,
      "step": 3430
    },
    {
      "epoch": 6.88,
      "grad_norm": 3.6166296005249023,
      "learning_rate": 5.41040287534517e-06,
      "loss": 0.4067,
      "step": 3440
    },
    {
      "epoch": 6.9,
      "grad_norm": 2.424628257751465,
      "learning_rate": 5.348489219225417e-06,
      "loss": 0.4424,
      "step": 3450
    },
    {
      "epoch": 6.92,
      "grad_norm": 2.7839276790618896,
      "learning_rate": 5.286802271178815e-06,
      "loss": 0.4508,
      "step": 3460
    },
    {
      "epoch": 6.9399999999999995,
      "grad_norm": 3.2447237968444824,
      "learning_rate": 5.225345037740186e-06,
      "loss": 0.3984,
      "step": 3470
    },
    {
      "epoch": 6.96,
      "grad_norm": 1.6200125217437744,
      "learning_rate": 5.16412051424839e-06,
      "loss": 0.4499,
      "step": 3480
    },
    {
      "epoch": 6.98,
      "grad_norm": 1.5760829448699951,
      "learning_rate": 5.103131684700315e-06,
      "loss": 0.4154,
      "step": 3490
    },
    {
      "epoch": 7.0,
      "grad_norm": 11.93620491027832,
      "learning_rate": 5.042381521605473e-06,
      "loss": 0.391,
      "step": 3500
    },
    {
      "epoch": 7.02,
      "grad_norm": 4.725837707519531,
      "learning_rate": 4.981872985841115e-06,
      "loss": 0.38,
      "step": 3510
    },
    {
      "epoch": 7.04,
      "grad_norm": 1.6126409769058228,
      "learning_rate": 4.921609026507907e-06,
      "loss": 0.3478,
      "step": 3520
    },
    {
      "epoch": 7.06,
      "grad_norm": 2.6732842922210693,
      "learning_rate": 4.861592580786205e-06,
      "loss": 0.3712,
      "step": 3530
    },
    {
      "epoch": 7.08,
      "grad_norm": 2.0151851177215576,
      "learning_rate": 4.801826573792905e-06,
      "loss": 0.3801,
      "step": 3540
    },
    {
      "epoch": 7.1,
      "grad_norm": 1.151302456855774,
      "learning_rate": 4.7423139184388725e-06,
      "loss": 0.3881,
      "step": 3550
    },
    {
      "epoch": 7.12,
      "grad_norm": 3.5267462730407715,
      "learning_rate": 4.6830575152869615e-06,
      "loss": 0.3489,
      "step": 3560
    },
    {
      "epoch": 7.14,
      "grad_norm": 1.8429639339447021,
      "learning_rate": 4.62406025241067e-06,
      "loss": 0.4284,
      "step": 3570
    },
    {
      "epoch": 7.16,
      "grad_norm": 4.2320051193237305,
      "learning_rate": 4.565325005253356e-06,
      "loss": 0.4055,
      "step": 3580
    },
    {
      "epoch": 7.18,
      "grad_norm": 2.368800163269043,
      "learning_rate": 4.506854636488103e-06,
      "loss": 0.3627,
      "step": 3590
    },
    {
      "epoch": 7.2,
      "grad_norm": 7.869661331176758,
      "learning_rate": 4.44865199587819e-06,
      "loss": 0.3866,
      "step": 3600
    },
    {
      "epoch": 7.22,
      "grad_norm": 1.2418557405471802,
      "learning_rate": 4.39071992013822e-06,
      "loss": 0.3947,
      "step": 3610
    },
    {
      "epoch": 7.24,
      "grad_norm": 1.7556277513504028,
      "learning_rate": 4.3330612327958265e-06,
      "loss": 0.4266,
      "step": 3620
    },
    {
      "epoch": 7.26,
      "grad_norm": 4.239712238311768,
      "learning_rate": 4.275678744054094e-06,
      "loss": 0.3495,
      "step": 3630
    },
    {
      "epoch": 7.28,
      "grad_norm": 2.917245626449585,
      "learning_rate": 4.218575250654559e-06,
      "loss": 0.4153,
      "step": 3640
    },
    {
      "epoch": 7.3,
      "grad_norm": 1.490869402885437,
      "learning_rate": 4.161753535740932e-06,
      "loss": 0.3819,
      "step": 3650
    },
    {
      "epoch": 7.32,
      "grad_norm": 1.5143734216690063,
      "learning_rate": 4.105216368723437e-06,
      "loss": 0.4032,
      "step": 3660
    },
    {
      "epoch": 7.34,
      "grad_norm": 3.434727907180786,
      "learning_rate": 4.048966505143831e-06,
      "loss": 0.358,
      "step": 3670
    },
    {
      "epoch": 7.36,
      "grad_norm": 1.666413426399231,
      "learning_rate": 3.993006686541108e-06,
      "loss": 0.4101,
      "step": 3680
    },
    {
      "epoch": 7.38,
      "grad_norm": 2.142817974090576,
      "learning_rate": 3.937339640317879e-06,
      "loss": 0.3803,
      "step": 3690
    },
    {
      "epoch": 7.4,
      "grad_norm": 0.9919471740722656,
      "learning_rate": 3.88196807960744e-06,
      "loss": 0.3844,
      "step": 3700
    },
    {
      "epoch": 7.42,
      "grad_norm": 2.370820999145508,
      "learning_rate": 3.826894703141552e-06,
      "loss": 0.3536,
      "step": 3710
    },
    {
      "epoch": 7.44,
      "grad_norm": 1.761391520500183,
      "learning_rate": 3.772122195118877e-06,
      "loss": 0.3957,
      "step": 3720
    },
    {
      "epoch": 7.46,
      "grad_norm": 1.3135240077972412,
      "learning_rate": 3.7176532250741857e-06,
      "loss": 0.4308,
      "step": 3730
    },
    {
      "epoch": 7.48,
      "grad_norm": 7.84911584854126,
      "learning_rate": 3.663490447748236e-06,
      "loss": 0.3988,
      "step": 3740
    },
    {
      "epoch": 7.5,
      "grad_norm": 2.369114875793457,
      "learning_rate": 3.6096365029583803e-06,
      "loss": 0.3983,
      "step": 3750
    },
    {
      "epoch": 7.52,
      "grad_norm": 3.6399729251861572,
      "learning_rate": 3.5560940154699133e-06,
      "loss": 0.37,
      "step": 3760
    },
    {
      "epoch": 7.54,
      "grad_norm": 3.580399751663208,
      "learning_rate": 3.502865594868136e-06,
      "loss": 0.3645,
      "step": 3770
    },
    {
      "epoch": 7.5600000000000005,
      "grad_norm": 3.4667141437530518,
      "learning_rate": 3.4499538354311757e-06,
      "loss": 0.4179,
      "step": 3780
    },
    {
      "epoch": 7.58,
      "grad_norm": 2.440298318862915,
      "learning_rate": 3.397361316003539e-06,
      "loss": 0.324,
      "step": 3790
    },
    {
      "epoch": 7.6,
      "grad_norm": 2.0638418197631836,
      "learning_rate": 3.3450905998704274e-06,
      "loss": 0.3789,
      "step": 3800
    },
    {
      "epoch": 7.62,
      "grad_norm": 2.4778010845184326,
      "learning_rate": 3.2931442346328e-06,
      "loss": 0.3608,
      "step": 3810
    },
    {
      "epoch": 7.64,
      "grad_norm": 2.4514052867889404,
      "learning_rate": 3.241524752083215e-06,
      "loss": 0.3985,
      "step": 3820
    },
    {
      "epoch": 7.66,
      "grad_norm": 1.3875998258590698,
      "learning_rate": 3.190234668082427e-06,
      "loss": 0.3447,
      "step": 3830
    },
    {
      "epoch": 7.68,
      "grad_norm": 1.2562263011932373,
      "learning_rate": 3.1392764824367706e-06,
      "loss": 0.3426,
      "step": 3840
    },
    {
      "epoch": 7.7,
      "grad_norm": 1.41866135597229,
      "learning_rate": 3.0886526787763237e-06,
      "loss": 0.3576,
      "step": 3850
    },
    {
      "epoch": 7.72,
      "grad_norm": 3.5626509189605713,
      "learning_rate": 3.038365724433858e-06,
      "loss": 0.3928,
      "step": 3860
    },
    {
      "epoch": 7.74,
      "grad_norm": 5.021074295043945,
      "learning_rate": 2.988418070324577e-06,
      "loss": 0.3589,
      "step": 3870
    },
    {
      "epoch": 7.76,
      "grad_norm": 1.392564058303833,
      "learning_rate": 2.938812150826684e-06,
      "loss": 0.3851,
      "step": 3880
    },
    {
      "epoch": 7.78,
      "grad_norm": 3.992396116256714,
      "learning_rate": 2.8895503836627105e-06,
      "loss": 0.3688,
      "step": 3890
    },
    {
      "epoch": 7.8,
      "grad_norm": 2.653534173965454,
      "learning_rate": 2.840635169781688e-06,
      "loss": 0.3585,
      "step": 3900
    },
    {
      "epoch": 7.82,
      "grad_norm": 6.615116596221924,
      "learning_rate": 2.7920688932421337e-06,
      "loss": 0.3653,
      "step": 3910
    },
    {
      "epoch": 7.84,
      "grad_norm": 5.369716644287109,
      "learning_rate": 2.7438539210958483e-06,
      "loss": 0.3512,
      "step": 3920
    },
    {
      "epoch": 7.86,
      "grad_norm": 1.679790735244751,
      "learning_rate": 2.6959926032725537e-06,
      "loss": 0.3717,
      "step": 3930
    },
    {
      "epoch": 7.88,
      "grad_norm": 2.233903646469116,
      "learning_rate": 2.648487272465361e-06,
      "loss": 0.3806,
      "step": 3940
    },
    {
      "epoch": 7.9,
      "grad_norm": 3.0028107166290283,
      "learning_rate": 2.6013402440170676e-06,
      "loss": 0.3993,
      "step": 3950
    },
    {
      "epoch": 7.92,
      "grad_norm": 3.600489854812622,
      "learning_rate": 2.5545538158073278e-06,
      "loss": 0.3387,
      "step": 3960
    },
    {
      "epoch": 7.9399999999999995,
      "grad_norm": 7.545295715332031,
      "learning_rate": 2.512756228659141e-06,
      "loss": 0.37,
      "step": 3970
    },
    {
      "epoch": 7.96,
      "grad_norm": 2.566960573196411,
      "learning_rate": 2.4666612085261344e-06,
      "loss": 0.3967,
      "step": 3980
    },
    {
      "epoch": 7.98,
      "grad_norm": 2.3997247219085693,
      "learning_rate": 2.420933352697865e-06,
      "loss": 0.4029,
      "step": 3990
    },
    {
      "epoch": 8.0,
      "grad_norm": 2.916670560836792,
      "learning_rate": 2.37557488988552e-06,
      "loss": 0.3713,
      "step": 4000
    },
    {
      "epoch": 8.02,
      "grad_norm": 1.7952624559402466,
      "learning_rate": 2.3305880307965834e-06,
      "loss": 0.3232,
      "step": 4010
    },
    {
      "epoch": 8.04,
      "grad_norm": 1.91434645652771,
      "learning_rate": 2.2859749680270983e-06,
      "loss": 0.331,
      "step": 4020
    },
    {
      "epoch": 8.06,
      "grad_norm": 3.671706438064575,
      "learning_rate": 2.241737875954808e-06,
      "loss": 0.3818,
      "step": 4030
    },
    {
      "epoch": 8.08,
      "grad_norm": 1.5308889150619507,
      "learning_rate": 2.1978789106331666e-06,
      "loss": 0.3482,
      "step": 4040
    },
    {
      "epoch": 8.1,
      "grad_norm": 1.8674166202545166,
      "learning_rate": 2.154400209686268e-06,
      "loss": 0.3195,
      "step": 4050
    },
    {
      "epoch": 8.12,
      "grad_norm": 1.5842407941818237,
      "learning_rate": 2.1113038922046603e-06,
      "loss": 0.3557,
      "step": 4060
    },
    {
      "epoch": 8.14,
      "grad_norm": 2.2769813537597656,
      "learning_rate": 2.0685920586420562e-06,
      "loss": 0.2853,
      "step": 4070
    },
    {
      "epoch": 8.16,
      "grad_norm": 1.7789726257324219,
      "learning_rate": 2.026266790712965e-06,
      "loss": 0.316,
      "step": 4080
    },
    {
      "epoch": 8.18,
      "grad_norm": 2.544579029083252,
      "learning_rate": 1.984330151291233e-06,
      "loss": 0.3328,
      "step": 4090
    },
    {
      "epoch": 8.2,
      "grad_norm": 5.644877910614014,
      "learning_rate": 1.9427841843095063e-06,
      "loss": 0.3338,
      "step": 4100
    },
    {
      "epoch": 8.22,
      "grad_norm": 2.9125707149505615,
      "learning_rate": 1.9016309146596024e-06,
      "loss": 0.3226,
      "step": 4110
    },
    {
      "epoch": 8.24,
      "grad_norm": 2.9386703968048096,
      "learning_rate": 1.8608723480938207e-06,
      "loss": 0.3147,
      "step": 4120
    },
    {
      "epoch": 8.26,
      "grad_norm": 5.057535648345947,
      "learning_rate": 1.820510471127196e-06,
      "loss": 0.3549,
      "step": 4130
    },
    {
      "epoch": 8.28,
      "grad_norm": 1.1568169593811035,
      "learning_rate": 1.7805472509406695e-06,
      "loss": 0.3701,
      "step": 4140
    },
    {
      "epoch": 8.3,
      "grad_norm": 2.978498697280884,
      "learning_rate": 1.7409846352852144e-06,
      "loss": 0.341,
      "step": 4150
    },
    {
      "epoch": 8.32,
      "grad_norm": 1.8623732328414917,
      "learning_rate": 1.7018245523869038e-06,
      "loss": 0.2754,
      "step": 4160
    },
    {
      "epoch": 8.34,
      "grad_norm": 3.46683406829834,
      "learning_rate": 1.6630689108529286e-06,
      "loss": 0.3958,
      "step": 4170
    },
    {
      "epoch": 8.36,
      "grad_norm": 5.015219211578369,
      "learning_rate": 1.6247195995785836e-06,
      "loss": 0.3512,
      "step": 4180
    },
    {
      "epoch": 8.38,
      "grad_norm": 1.5242810249328613,
      "learning_rate": 1.5867784876551973e-06,
      "loss": 0.3533,
      "step": 4190
    },
    {
      "epoch": 8.4,
      "grad_norm": 4.693676948547363,
      "learning_rate": 1.5492474242790368e-06,
      "loss": 0.3746,
      "step": 4200
    },
    {
      "epoch": 8.42,
      "grad_norm": 2.436262845993042,
      "learning_rate": 1.5121282386611823e-06,
      "loss": 0.3274,
      "step": 4210
    },
    {
      "epoch": 8.44,
      "grad_norm": 2.2660608291625977,
      "learning_rate": 1.4754227399383758e-06,
      "loss": 0.3055,
      "step": 4220
    },
    {
      "epoch": 8.46,
      "grad_norm": 2.7948834896087646,
      "learning_rate": 1.439132717084839e-06,
      "loss": 0.3078,
      "step": 4230
    },
    {
      "epoch": 8.48,
      "grad_norm": 1.3765865564346313,
      "learning_rate": 1.40325993882509e-06,
      "loss": 0.3194,
      "step": 4240
    },
    {
      "epoch": 8.5,
      "grad_norm": 1.2223212718963623,
      "learning_rate": 1.3678061535477305e-06,
      "loss": 0.352,
      "step": 4250
    },
    {
      "epoch": 8.52,
      "grad_norm": 2.556001663208008,
      "learning_rate": 1.3327730892202384e-06,
      "loss": 0.3061,
      "step": 4260
    },
    {
      "epoch": 8.54,
      "grad_norm": 4.0893168449401855,
      "learning_rate": 1.2981624533047432e-06,
      "loss": 0.406,
      "step": 4270
    },
    {
      "epoch": 8.56,
      "grad_norm": 1.8102929592132568,
      "learning_rate": 1.2639759326748136e-06,
      "loss": 0.3335,
      "step": 4280
    },
    {
      "epoch": 8.58,
      "grad_norm": 0.6934239268302917,
      "learning_rate": 1.230215193533233e-06,
      "loss": 0.4048,
      "step": 4290
    },
    {
      "epoch": 8.6,
      "grad_norm": 2.5495901107788086,
      "learning_rate": 1.196881881330798e-06,
      "loss": 0.3388,
      "step": 4300
    },
    {
      "epoch": 8.62,
      "grad_norm": 2.681366443634033,
      "learning_rate": 1.1639776206861197e-06,
      "loss": 0.358,
      "step": 4310
    },
    {
      "epoch": 8.64,
      "grad_norm": 1.624990463256836,
      "learning_rate": 1.1315040153064416e-06,
      "loss": 0.3628,
      "step": 4320
    },
    {
      "epoch": 8.66,
      "grad_norm": 7.331467151641846,
      "learning_rate": 1.0994626479094749e-06,
      "loss": 0.3585,
      "step": 4330
    },
    {
      "epoch": 8.68,
      "grad_norm": 1.2213658094406128,
      "learning_rate": 1.0678550801462662e-06,
      "loss": 0.3583,
      "step": 4340
    },
    {
      "epoch": 8.7,
      "grad_norm": 2.539713144302368,
      "learning_rate": 1.0366828525250728e-06,
      "loss": 0.2861,
      "step": 4350
    },
    {
      "epoch": 8.72,
      "grad_norm": 4.281270980834961,
      "learning_rate": 1.0059474843362893e-06,
      "loss": 0.3422,
      "step": 4360
    },
    {
      "epoch": 8.74,
      "grad_norm": 1.515568733215332,
      "learning_rate": 9.756504735784067e-07,
      "loss": 0.3337,
      "step": 4370
    },
    {
      "epoch": 8.76,
      "grad_norm": 1.730093240737915,
      "learning_rate": 9.457932968849826e-07,
      "loss": 0.3163,
      "step": 4380
    },
    {
      "epoch": 8.78,
      "grad_norm": 4.305525302886963,
      "learning_rate": 9.16377409452689e-07,
      "loss": 0.3132,
      "step": 4390
    },
    {
      "epoch": 8.8,
      "grad_norm": 1.6857116222381592,
      "learning_rate": 8.874042449703779e-07,
      "loss": 0.3108,
      "step": 4400
    },
    {
      "epoch": 8.82,
      "grad_norm": 1.7638370990753174,
      "learning_rate": 8.58875215549212e-07,
      "loss": 0.3444,
      "step": 4410
    },
    {
      "epoch": 8.84,
      "grad_norm": 2.9410483837127686,
      "learning_rate": 8.307917116538378e-07,
      "loss": 0.3582,
      "step": 4420
    },
    {
      "epoch": 8.86,
      "grad_norm": 1.4133245944976807,
      "learning_rate": 8.031551020346129e-07,
      "loss": 0.3014,
      "step": 4430
    },
    {
      "epoch": 8.88,
      "grad_norm": 2.466925621032715,
      "learning_rate": 7.759667336609011e-07,
      "loss": 0.3578,
      "step": 4440
    },
    {
      "epoch": 8.9,
      "grad_norm": 1.979108214378357,
      "learning_rate": 7.492279316554207e-07,
      "loss": 0.3253,
      "step": 4450
    },
    {
      "epoch": 8.92,
      "grad_norm": 1.6241674423217773,
      "learning_rate": 7.22939999229657e-07,
      "loss": 0.3839,
      "step": 4460
    },
    {
      "epoch": 8.94,
      "grad_norm": 2.58152174949646,
      "learning_rate": 6.971042176203535e-07,
      "loss": 0.268,
      "step": 4470
    },
    {
      "epoch": 8.96,
      "grad_norm": 2.5680618286132812,
      "learning_rate": 6.717218460270536e-07,
      "loss": 0.332,
      "step": 4480
    },
    {
      "epoch": 8.98,
      "grad_norm": 2.1524252891540527,
      "learning_rate": 6.467941215507434e-07,
      "loss": 0.361,
      "step": 4490
    },
    {
      "epoch": 9.0,
      "grad_norm": 1.6696678400039673,
      "learning_rate": 6.223222591335409e-07,
      "loss": 0.3358,
      "step": 4500
    },
    {
      "epoch": 9.02,
      "grad_norm": 1.1763229370117188,
      "learning_rate": 5.98307451499498e-07,
      "loss": 0.2874,
      "step": 4510
    },
    {
      "epoch": 9.04,
      "grad_norm": 1.381579041481018,
      "learning_rate": 5.747508690964599e-07,
      "loss": 0.361,
      "step": 4520
    },
    {
      "epoch": 9.06,
      "grad_norm": 1.560950756072998,
      "learning_rate": 5.516536600390188e-07,
      "loss": 0.2929,
      "step": 4530
    },
    {
      "epoch": 9.08,
      "grad_norm": 2.703350782394409,
      "learning_rate": 5.290169500525577e-07,
      "loss": 0.2854,
      "step": 4540
    },
    {
      "epoch": 9.1,
      "grad_norm": 1.5632970333099365,
      "learning_rate": 5.068418424183874e-07,
      "loss": 0.3173,
      "step": 4550
    },
    {
      "epoch": 9.12,
      "grad_norm": 1.8101422786712646,
      "learning_rate": 4.851294179199673e-07,
      "loss": 0.3683,
      "step": 4560
    },
    {
      "epoch": 9.14,
      "grad_norm": 1.0653437376022339,
      "learning_rate": 4.638807347902408e-07,
      "loss": 0.3256,
      "step": 4570
    },
    {
      "epoch": 9.16,
      "grad_norm": 2.522818088531494,
      "learning_rate": 4.4309682866004124e-07,
      "loss": 0.319,
      "step": 4580
    },
    {
      "epoch": 9.18,
      "grad_norm": 3.289451837539673,
      "learning_rate": 4.2277871250763327e-07,
      "loss": 0.3221,
      "step": 4590
    },
    {
      "epoch": 9.2,
      "grad_norm": 2.0382297039031982,
      "learning_rate": 4.0292737660933335e-07,
      "loss": 0.2951,
      "step": 4600
    },
    {
      "epoch": 9.22,
      "grad_norm": 2.1435165405273438,
      "learning_rate": 3.835437884912474e-07,
      "loss": 0.3738,
      "step": 4610
    },
    {
      "epoch": 9.24,
      "grad_norm": 1.6461173295974731,
      "learning_rate": 3.646288928821151e-07,
      "loss": 0.2898,
      "step": 4620
    },
    {
      "epoch": 9.26,
      "grad_norm": 2.4130804538726807,
      "learning_rate": 3.4618361166726123e-07,
      "loss": 0.3792,
      "step": 4630
    },
    {
      "epoch": 9.28,
      "grad_norm": 2.4446017742156982,
      "learning_rate": 3.282088438436715e-07,
      "loss": 0.3424,
      "step": 4640
    },
    {
      "epoch": 9.3,
      "grad_norm": 1.3320859670639038,
      "learning_rate": 3.10705465476171e-07,
      "loss": 0.358,
      "step": 4650
    },
    {
      "epoch": 9.32,
      "grad_norm": 3.8511667251586914,
      "learning_rate": 2.936743296547273e-07,
      "loss": 0.32,
      "step": 4660
    },
    {
      "epoch": 9.34,
      "grad_norm": 5.630286693572998,
      "learning_rate": 2.771162664528726e-07,
      "loss": 0.3079,
      "step": 4670
    },
    {
      "epoch": 9.36,
      "grad_norm": 1.600219964981079,
      "learning_rate": 2.6103208288724815e-07,
      "loss": 0.2834,
      "step": 4680
    },
    {
      "epoch": 9.38,
      "grad_norm": 1.0820380449295044,
      "learning_rate": 2.4542256287826915e-07,
      "loss": 0.354,
      "step": 4690
    },
    {
      "epoch": 9.4,
      "grad_norm": 1.4870035648345947,
      "learning_rate": 2.3028846721191878e-07,
      "loss": 0.3243,
      "step": 4700
    },
    {
      "epoch": 9.42,
      "grad_norm": 3.148569107055664,
      "learning_rate": 2.1563053350266983e-07,
      "loss": 0.3121,
      "step": 4710
    },
    {
      "epoch": 9.44,
      "grad_norm": 1.8829195499420166,
      "learning_rate": 2.014494761575314e-07,
      "loss": 0.3142,
      "step": 4720
    },
    {
      "epoch": 9.46,
      "grad_norm": 3.1038215160369873,
      "learning_rate": 1.877459863412323e-07,
      "loss": 0.3287,
      "step": 4730
    },
    {
      "epoch": 9.48,
      "grad_norm": 1.9286001920700073,
      "learning_rate": 1.7452073194253237e-07,
      "loss": 0.2989,
      "step": 4740
    },
    {
      "epoch": 9.5,
      "grad_norm": 2.0495471954345703,
      "learning_rate": 1.6177435754167413e-07,
      "loss": 0.3632,
      "step": 4750
    },
    {
      "epoch": 9.52,
      "grad_norm": 2.1833696365356445,
      "learning_rate": 1.4950748437896235e-07,
      "loss": 0.265,
      "step": 4760
    },
    {
      "epoch": 9.54,
      "grad_norm": 3.15493106842041,
      "learning_rate": 1.377207103244904e-07,
      "loss": 0.283,
      "step": 4770
    },
    {
      "epoch": 9.56,
      "grad_norm": 1.2273836135864258,
      "learning_rate": 1.26414609848996e-07,
      "loss": 0.2264,
      "step": 4780
    },
    {
      "epoch": 9.58,
      "grad_norm": 1.6316149234771729,
      "learning_rate": 1.1558973399586671e-07,
      "loss": 0.3198,
      "step": 4790
    },
    {
      "epoch": 9.6,
      "grad_norm": 2.2700629234313965,
      "learning_rate": 1.052466103542793e-07,
      "loss": 0.2258,
      "step": 4800
    },
    {
      "epoch": 9.62,
      "grad_norm": 1.132501244544983,
      "learning_rate": 9.538574303348813e-08,
      "loss": 0.3053,
      "step": 4810
    },
    {
      "epoch": 9.64,
      "grad_norm": 1.9846259355545044,
      "learning_rate": 8.600761263825475e-08,
      "loss": 0.278,
      "step": 4820
    },
    {
      "epoch": 9.66,
      "grad_norm": 1.3777711391448975,
      "learning_rate": 7.71126762454233e-08,
      "loss": 0.3211,
      "step": 4830
    },
    {
      "epoch": 9.68,
      "grad_norm": 1.355865240097046,
      "learning_rate": 6.870136738164612e-08,
      "loss": 0.3079,
      "step": 4840
    },
    {
      "epoch": 9.7,
      "grad_norm": 2.8824238777160645,
      "learning_rate": 6.07740960022507e-08,
      "loss": 0.3717,
      "step": 4850
    },
    {
      "epoch": 9.72,
      "grad_norm": 2.4788002967834473,
      "learning_rate": 5.3331248471258926e-08,
      "loss": 0.3052,
      "step": 4860
    },
    {
      "epoch": 9.74,
      "grad_norm": 1.3350954055786133,
      "learning_rate": 4.6373187542561036e-08,
      "loss": 0.3018,
      "step": 4870
    },
    {
      "epoch": 9.76,
      "grad_norm": 1.485877513885498,
      "learning_rate": 3.990025234222872e-08,
      "loss": 0.2694,
      "step": 4880
    },
    {
      "epoch": 9.78,
      "grad_norm": 2.1320180892944336,
      "learning_rate": 3.391275835199159e-08,
      "loss": 0.323,
      "step": 4890
    },
    {
      "epoch": 9.8,
      "grad_norm": 1.2034118175506592,
      "learning_rate": 2.8410997393860663e-08,
      "loss": 0.302,
      "step": 4900
    },
    {
      "epoch": 9.82,
      "grad_norm": 1.9911288022994995,
      "learning_rate": 2.339523761590301e-08,
      "loss": 0.3561,
      "step": 4910
    },
    {
      "epoch": 9.84,
      "grad_norm": 4.588063716888428,
      "learning_rate": 1.886572347917337e-08,
      "loss": 0.3486,
      "step": 4920
    },
    {
      "epoch": 9.86,
      "grad_norm": 1.2661594152450562,
      "learning_rate": 1.482267574580143e-08,
      "loss": 0.3651,
      "step": 4930
    },
    {
      "epoch": 9.88,
      "grad_norm": 1.7865337133407593,
      "learning_rate": 1.126629146822933e-08,
      "loss": 0.2544,
      "step": 4940
    },
    {
      "epoch": 9.9,
      "grad_norm": 3.7908692359924316,
      "learning_rate": 8.196743979610455e-09,
      "loss": 0.2575,
      "step": 4950
    },
    {
      "epoch": 9.92,
      "grad_norm": 2.3560678958892822,
      "learning_rate": 5.614182885357311e-09,
      "loss": 0.2833,
      "step": 4960
    },
    {
      "epoch": 9.94,
      "grad_norm": 1.629074215888977,
      "learning_rate": 3.518734055855122e-09,
      "loss": 0.3345,
      "step": 4970
    },
    {
      "epoch": 9.96,
      "grad_norm": 2.738832473754883,
      "learning_rate": 1.910499620322304e-09,
      "loss": 0.3034,
      "step": 4980
    },
    {
      "epoch": 9.98,
      "grad_norm": 1.9189926385879517,
      "learning_rate": 7.895579618388827e-10,
      "loss": 0.2543,
      "step": 4990
    },
    {
      "epoch": 10.0,
      "grad_norm": 1.597418189048767,
      "learning_rate": 1.559637135173375e-10,
      "loss": 0.3149,
      "step": 5000
    },
    {
      "epoch": 10.0,
      "step": 5000,
      "total_flos": 1.1899956660640154e+17,
      "train_loss": 0.49240388979911803,
      "train_runtime": 35927.6647,
      "train_samples_per_second": 0.278,
      "train_steps_per_second": 0.139
    }
  ],
  "logging_steps": 10,
  "max_steps": 5000,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 100,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.1899956660640154e+17,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}