{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 1.9255575164447296,
  "global_step": 4000,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 1.8e-07,
      "loss": 0.6934,
      "step": 10
    },
    {
      "epoch": 0.01,
      "learning_rate": 3.8e-07,
      "loss": 0.6932,
      "step": 20
    },
    {
      "epoch": 0.01,
      "learning_rate": 5.800000000000001e-07,
      "loss": 0.6941,
      "step": 30
    },
    {
      "epoch": 0.02,
      "learning_rate": 7.8e-07,
      "loss": 0.6938,
      "step": 40
    },
    {
      "epoch": 0.02,
      "learning_rate": 9.800000000000001e-07,
      "loss": 0.693,
      "step": 50
    },
    {
      "epoch": 0.03,
      "learning_rate": 1.1800000000000001e-06,
      "loss": 0.6943,
      "step": 60
    },
    {
      "epoch": 0.03,
      "learning_rate": 1.3800000000000001e-06,
      "loss": 0.6945,
      "step": 70
    },
    {
      "epoch": 0.04,
      "learning_rate": 1.5800000000000001e-06,
      "loss": 0.693,
      "step": 80
    },
    {
      "epoch": 0.04,
      "learning_rate": 1.7800000000000001e-06,
      "loss": 0.6935,
      "step": 90
    },
    {
      "epoch": 0.05,
      "learning_rate": 1.98e-06,
      "loss": 0.6937,
      "step": 100
    },
    {
      "epoch": 0.05,
      "learning_rate": 2.1800000000000003e-06,
      "loss": 0.6932,
      "step": 110
    },
    {
      "epoch": 0.06,
      "learning_rate": 2.38e-06,
      "loss": 0.6936,
      "step": 120
    },
    {
      "epoch": 0.06,
      "learning_rate": 2.5800000000000003e-06,
      "loss": 0.6932,
      "step": 130
    },
    {
      "epoch": 0.07,
      "learning_rate": 2.7800000000000005e-06,
      "loss": 0.6938,
      "step": 140
    },
    {
      "epoch": 0.07,
      "learning_rate": 2.9800000000000003e-06,
      "loss": 0.6937,
      "step": 150
    },
    {
      "epoch": 0.08,
      "learning_rate": 3.1800000000000005e-06,
      "loss": 0.6925,
      "step": 160
    },
    {
      "epoch": 0.08,
      "learning_rate": 3.3800000000000007e-06,
      "loss": 0.6923,
      "step": 170
    },
    {
      "epoch": 0.09,
      "learning_rate": 3.58e-06,
      "loss": 0.6928,
      "step": 180
    },
    {
      "epoch": 0.09,
      "learning_rate": 3.7800000000000002e-06,
      "loss": 0.6899,
      "step": 190
    },
    {
      "epoch": 0.1,
      "learning_rate": 3.980000000000001e-06,
      "loss": 0.691,
      "step": 200
    },
    {
      "epoch": 0.1,
      "learning_rate": 4.18e-06,
      "loss": 0.6897,
      "step": 210
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.38e-06,
      "loss": 0.6874,
      "step": 220
    },
    {
      "epoch": 0.11,
      "learning_rate": 4.58e-06,
      "loss": 0.6784,
      "step": 230
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.78e-06,
      "loss": 0.6619,
      "step": 240
    },
    {
      "epoch": 0.12,
      "learning_rate": 4.980000000000001e-06,
      "loss": 0.6517,
      "step": 250
    },
    {
      "epoch": 0.13,
      "learning_rate": 5.18e-06,
      "loss": 0.6444,
      "step": 260
    },
    {
      "epoch": 0.13,
      "learning_rate": 5.380000000000001e-06,
      "loss": 0.6422,
      "step": 270
    },
    {
      "epoch": 0.13,
      "learning_rate": 5.580000000000001e-06,
      "loss": 0.6389,
      "step": 280
    },
    {
      "epoch": 0.14,
      "learning_rate": 5.78e-06,
      "loss": 0.6255,
      "step": 290
    },
    {
      "epoch": 0.14,
      "learning_rate": 5.98e-06,
      "loss": 0.6281,
      "step": 300
    },
    {
      "epoch": 0.15,
      "learning_rate": 6.18e-06,
      "loss": 0.6368,
      "step": 310
    },
    {
      "epoch": 0.15,
      "learning_rate": 6.380000000000001e-06,
      "loss": 0.6291,
      "step": 320
    },
    {
      "epoch": 0.16,
      "learning_rate": 6.5800000000000005e-06,
      "loss": 0.6103,
      "step": 330
    },
    {
      "epoch": 0.16,
      "learning_rate": 6.780000000000001e-06,
      "loss": 0.611,
      "step": 340
    },
    {
      "epoch": 0.17,
      "learning_rate": 6.98e-06,
      "loss": 0.6003,
      "step": 350
    },
    {
      "epoch": 0.17,
      "learning_rate": 7.180000000000001e-06,
      "loss": 0.607,
      "step": 360
    },
    {
      "epoch": 0.18,
      "learning_rate": 7.3800000000000005e-06,
      "loss": 0.6062,
      "step": 370
    },
    {
      "epoch": 0.18,
      "learning_rate": 7.5600000000000005e-06,
      "loss": 0.6014,
      "step": 380
    },
    {
      "epoch": 0.19,
      "learning_rate": 7.76e-06,
      "loss": 0.6156,
      "step": 390
    },
    {
      "epoch": 0.19,
      "learning_rate": 7.960000000000002e-06,
      "loss": 0.6034,
      "step": 400
    },
    {
      "epoch": 0.2,
      "learning_rate": 8.16e-06,
      "loss": 0.5974,
      "step": 410
    },
    {
      "epoch": 0.2,
      "learning_rate": 8.36e-06,
      "loss": 0.6065,
      "step": 420
    },
    {
      "epoch": 0.21,
      "learning_rate": 8.560000000000001e-06,
      "loss": 0.5785,
      "step": 430
    },
    {
      "epoch": 0.21,
      "learning_rate": 8.76e-06,
      "loss": 0.601,
      "step": 440
    },
    {
      "epoch": 0.22,
      "learning_rate": 8.96e-06,
      "loss": 0.5838,
      "step": 450
    },
    {
      "epoch": 0.22,
      "learning_rate": 9.16e-06,
      "loss": 0.6045,
      "step": 460
    },
    {
      "epoch": 0.23,
      "learning_rate": 9.360000000000002e-06,
      "loss": 0.5914,
      "step": 470
    },
    {
      "epoch": 0.23,
      "learning_rate": 9.56e-06,
      "loss": 0.585,
      "step": 480
    },
    {
      "epoch": 0.24,
      "learning_rate": 9.760000000000001e-06,
      "loss": 0.5845,
      "step": 490
    },
    {
      "epoch": 0.24,
      "learning_rate": 9.960000000000001e-06,
      "loss": 0.5926,
      "step": 500
    },
    {
      "epoch": 0.24,
      "eval_webgpt_accuracy": 0.5497957099080695,
      "eval_webgpt_loss": 0.7101226449012756,
      "eval_webgpt_runtime": 39.1748,
      "eval_webgpt_samples_per_second": 99.962,
      "eval_webgpt_steps_per_second": 1.327,
      "step": 500
    },
    {
      "epoch": 0.24,
      "eval_hfsummary_accuracy": 0.6480669830426503,
      "eval_hfsummary_loss": 0.6175686717033386,
      "eval_hfsummary_runtime": 973.4644,
      "eval_hfsummary_samples_per_second": 33.985,
      "eval_hfsummary_steps_per_second": 0.448,
      "step": 500
    },
    {
      "epoch": 0.24,
      "eval_anthropic_rlhf_accuracy": 0.5922591206735267,
      "eval_anthropic_rlhf_loss": 0.6707426905632019,
      "eval_anthropic_rlhf_runtime": 90.7482,
      "eval_anthropic_rlhf_samples_per_second": 94.239,
      "eval_anthropic_rlhf_steps_per_second": 1.245,
      "step": 500
    },
    {
      "epoch": 0.24,
      "eval_gptsynthetic_accuracy": 0.9969834087481146,
      "eval_gptsynthetic_loss": 0.03679310530424118,
      "eval_gptsynthetic_runtime": 31.6689,
      "eval_gptsynthetic_samples_per_second": 104.677,
      "eval_gptsynthetic_steps_per_second": 1.389,
      "step": 500
    },
    {
      "epoch": 0.25,
      "learning_rate": 9.999974108670129e-06,
      "loss": 0.582,
      "step": 510
    },
    {
      "epoch": 0.25,
      "learning_rate": 9.999868925602092e-06,
      "loss": 0.5802,
      "step": 520
    },
    {
      "epoch": 0.26,
      "learning_rate": 9.99968283428854e-06,
      "loss": 0.5811,
      "step": 530
    },
    {
      "epoch": 0.26,
      "learning_rate": 9.99941583774082e-06,
      "loss": 0.5818,
      "step": 540
    },
    {
      "epoch": 0.26,
      "learning_rate": 9.999067940279488e-06,
      "loss": 0.5585,
      "step": 550
    },
    {
      "epoch": 0.27,
      "learning_rate": 9.998639147534247e-06,
      "loss": 0.5645,
      "step": 560
    },
    {
      "epoch": 0.27,
      "learning_rate": 9.998129466443849e-06,
      "loss": 0.584,
      "step": 570
    },
    {
      "epoch": 0.28,
      "learning_rate": 9.997538905255994e-06,
      "loss": 0.577,
      "step": 580
    },
    {
      "epoch": 0.28,
      "learning_rate": 9.996867473527188e-06,
      "loss": 0.5629,
      "step": 590
    },
    {
      "epoch": 0.29,
      "learning_rate": 9.996115182122585e-06,
      "loss": 0.5721,
      "step": 600
    },
    {
      "epoch": 0.29,
      "learning_rate": 9.995282043215824e-06,
      "loss": 0.5599,
      "step": 610
    },
    {
      "epoch": 0.3,
      "learning_rate": 9.994368070288814e-06,
      "loss": 0.5631,
      "step": 620
    },
    {
      "epoch": 0.3,
      "learning_rate": 9.993373278131538e-06,
      "loss": 0.5652,
      "step": 630
    },
    {
      "epoch": 0.31,
      "learning_rate": 9.99229768284179e-06,
      "loss": 0.557,
      "step": 640
    },
    {
      "epoch": 0.31,
      "learning_rate": 9.991141301824936e-06,
      "loss": 0.554,
      "step": 650
    },
    {
      "epoch": 0.32,
      "learning_rate": 9.989904153793614e-06,
      "loss": 0.5763,
      "step": 660
    },
    {
      "epoch": 0.32,
      "learning_rate": 9.98858625876745e-06,
      "loss": 0.5628,
      "step": 670
    },
    {
      "epoch": 0.33,
      "learning_rate": 9.987187638072715e-06,
      "loss": 0.5632,
      "step": 680
    },
    {
      "epoch": 0.33,
      "learning_rate": 9.98570831434199e-06,
      "loss": 0.5896,
      "step": 690
    },
    {
      "epoch": 0.34,
      "learning_rate": 9.984148311513805e-06,
      "loss": 0.5666,
      "step": 700
    },
    {
      "epoch": 0.34,
      "learning_rate": 9.982507654832232e-06,
      "loss": 0.5489,
      "step": 710
    },
    {
      "epoch": 0.35,
      "learning_rate": 9.980786370846504e-06,
      "loss": 0.5392,
      "step": 720
    },
    {
      "epoch": 0.35,
      "learning_rate": 9.978984487410564e-06,
      "loss": 0.5703,
      "step": 730
    },
    {
      "epoch": 0.36,
      "learning_rate": 9.977102033682621e-06,
      "loss": 0.5619,
      "step": 740
    },
    {
      "epoch": 0.36,
      "learning_rate": 9.97513904012468e-06,
      "loss": 0.5554,
      "step": 750
    },
    {
      "epoch": 0.37,
      "learning_rate": 9.973095538502044e-06,
      "loss": 0.5583,
      "step": 760
    },
    {
      "epoch": 0.37,
      "learning_rate": 9.97097156188281e-06,
      "loss": 0.5588,
      "step": 770
    },
    {
      "epoch": 0.38,
      "learning_rate": 9.96876714463732e-06,
      "loss": 0.5786,
      "step": 780
    },
    {
      "epoch": 0.38,
      "learning_rate": 9.966482322437616e-06,
      "loss": 0.5515,
      "step": 790
    },
    {
      "epoch": 0.39,
      "learning_rate": 9.96411713225686e-06,
      "loss": 0.5532,
      "step": 800
    },
    {
      "epoch": 0.39,
      "learning_rate": 9.961671612368731e-06,
      "loss": 0.5526,
      "step": 810
    },
    {
      "epoch": 0.39,
      "learning_rate": 9.959145802346815e-06,
      "loss": 0.5495,
      "step": 820
    },
    {
      "epoch": 0.4,
      "learning_rate": 9.95653974306395e-06,
      "loss": 0.5531,
      "step": 830
    },
    {
      "epoch": 0.4,
      "learning_rate": 9.95385347669158e-06,
      "loss": 0.5465,
      "step": 840
    },
    {
      "epoch": 0.41,
      "learning_rate": 9.951087046699066e-06,
      "loss": 0.5606,
      "step": 850
    },
    {
      "epoch": 0.41,
      "learning_rate": 9.94824049785298e-06,
      "loss": 0.546,
      "step": 860
    },
    {
      "epoch": 0.42,
      "learning_rate": 9.945313876216383e-06,
      "loss": 0.5503,
      "step": 870
    },
    {
      "epoch": 0.42,
      "learning_rate": 9.942307229148083e-06,
      "loss": 0.5395,
      "step": 880
    },
    {
      "epoch": 0.43,
      "learning_rate": 9.939220605301861e-06,
      "loss": 0.5466,
      "step": 890
    },
    {
      "epoch": 0.43,
      "learning_rate": 9.936054054625692e-06,
      "loss": 0.5672,
      "step": 900
    },
    {
      "epoch": 0.44,
      "learning_rate": 9.932807628360933e-06,
      "loss": 0.5491,
      "step": 910
    },
    {
      "epoch": 0.44,
      "learning_rate": 9.929481379041486e-06,
      "loss": 0.5549,
      "step": 920
    },
    {
      "epoch": 0.45,
      "learning_rate": 9.926075360492967e-06,
      "loss": 0.544,
      "step": 930
    },
    {
      "epoch": 0.45,
      "learning_rate": 9.922589627831814e-06,
      "loss": 0.5427,
      "step": 940
    },
    {
      "epoch": 0.46,
      "learning_rate": 9.919024237464411e-06,
      "loss": 0.5515,
      "step": 950
    },
    {
      "epoch": 0.46,
      "learning_rate": 9.915379247086166e-06,
      "loss": 0.5402,
      "step": 960
    },
    {
      "epoch": 0.47,
      "learning_rate": 9.911654715680578e-06,
      "loss": 0.5262,
      "step": 970
    },
    {
      "epoch": 0.47,
      "learning_rate": 9.907850703518292e-06,
      "loss": 0.5327,
      "step": 980
    },
    {
      "epoch": 0.48,
      "learning_rate": 9.903967272156108e-06,
      "loss": 0.5355,
      "step": 990
    },
    {
      "epoch": 0.48,
      "learning_rate": 9.900004484435997e-06,
      "loss": 0.5385,
      "step": 1000
    },
    {
      "epoch": 0.48,
      "eval_webgpt_accuracy": 0.5771195097037793,
      "eval_webgpt_loss": 0.6884612441062927,
      "eval_webgpt_runtime": 38.7402,
      "eval_webgpt_samples_per_second": 101.084,
      "eval_webgpt_steps_per_second": 1.342,
      "step": 1000
    },
    {
      "epoch": 0.48,
      "eval_hfsummary_accuracy": 0.6676843091618052,
      "eval_hfsummary_loss": 0.62443608045578,
      "eval_hfsummary_runtime": 978.4924,
      "eval_hfsummary_samples_per_second": 33.81,
      "eval_hfsummary_steps_per_second": 0.446,
      "step": 1000
    },
    {
      "epoch": 0.48,
      "eval_anthropic_rlhf_accuracy": 0.6214920486435921,
      "eval_anthropic_rlhf_loss": 0.6546406149864197,
      "eval_anthropic_rlhf_runtime": 91.0623,
      "eval_anthropic_rlhf_samples_per_second": 93.914,
      "eval_anthropic_rlhf_steps_per_second": 1.241,
      "step": 1000
    },
    {
      "epoch": 0.48,
      "eval_gptsynthetic_accuracy": 0.9963800904977376,
      "eval_gptsynthetic_loss": 0.019633520394563675,
      "eval_gptsynthetic_runtime": 32.5041,
      "eval_gptsynthetic_samples_per_second": 101.987,
      "eval_gptsynthetic_steps_per_second": 1.354,
      "step": 1000
    },
    {
      "epoch": 0.49,
      "learning_rate": 9.895962404484083e-06,
      "loss": 0.5368,
      "step": 1010
    },
    {
      "epoch": 0.49,
      "learning_rate": 9.891841097709599e-06,
      "loss": 0.5293,
      "step": 1020
    },
    {
      "epoch": 0.5,
      "learning_rate": 9.88764063080383e-06,
      "loss": 0.5373,
      "step": 1030
    },
    {
      "epoch": 0.5,
      "learning_rate": 9.883361071739045e-06,
      "loss": 0.5253,
      "step": 1040
    },
    {
      "epoch": 0.51,
      "learning_rate": 9.87900248976738e-06,
      "loss": 0.5406,
      "step": 1050
    },
    {
      "epoch": 0.51,
      "learning_rate": 9.874564955419725e-06,
      "loss": 0.5194,
      "step": 1060
    },
    {
      "epoch": 0.52,
      "learning_rate": 9.870048540504592e-06,
      "loss": 0.533,
      "step": 1070
    },
    {
      "epoch": 0.52,
      "learning_rate": 9.865453318106934e-06,
      "loss": 0.5138,
      "step": 1080
    },
    {
      "epoch": 0.52,
      "learning_rate": 9.860779362586978e-06,
      "loss": 0.5455,
      "step": 1090
    },
    {
      "epoch": 0.53,
      "learning_rate": 9.856026749579014e-06,
      "loss": 0.5369,
      "step": 1100
    },
    {
      "epoch": 0.53,
      "learning_rate": 9.851195555990178e-06,
      "loss": 0.5299,
      "step": 1110
    },
    {
      "epoch": 0.54,
      "learning_rate": 9.846285859999194e-06,
      "loss": 0.5544,
      "step": 1120
    },
    {
      "epoch": 0.54,
      "learning_rate": 9.841297741055124e-06,
      "loss": 0.5408,
      "step": 1130
    },
    {
      "epoch": 0.55,
      "learning_rate": 9.836231279876076e-06,
      "loss": 0.52,
      "step": 1140
    },
    {
      "epoch": 0.55,
      "learning_rate": 9.831086558447895e-06,
      "loss": 0.5383,
      "step": 1150
    },
    {
      "epoch": 0.56,
      "learning_rate": 9.82586366002284e-06,
      "loss": 0.5239,
      "step": 1160
    },
    {
      "epoch": 0.56,
      "learning_rate": 9.820562669118237e-06,
      "loss": 0.5239,
      "step": 1170
    },
    {
      "epoch": 0.57,
      "learning_rate": 9.815183671515107e-06,
      "loss": 0.5248,
      "step": 1180
    },
    {
      "epoch": 0.57,
      "learning_rate": 9.809726754256788e-06,
      "loss": 0.5193,
      "step": 1190
    },
    {
      "epoch": 0.58,
      "learning_rate": 9.804192005647512e-06,
      "loss": 0.5258,
      "step": 1200
    },
    {
      "epoch": 0.58,
      "learning_rate": 9.798579515250988e-06,
      "loss": 0.5361,
      "step": 1210
    },
    {
      "epoch": 0.59,
      "learning_rate": 9.792889373888946e-06,
      "loss": 0.5342,
      "step": 1220
    },
    {
      "epoch": 0.59,
      "learning_rate": 9.78712167363967e-06,
      "loss": 0.5263,
      "step": 1230
    },
    {
      "epoch": 0.6,
      "learning_rate": 9.781276507836513e-06,
      "loss": 0.5339,
      "step": 1240
    },
    {
      "epoch": 0.6,
      "learning_rate": 9.77535397106637e-06,
      "loss": 0.5457,
      "step": 1250
    },
    {
      "epoch": 0.61,
      "learning_rate": 9.76935415916817e-06,
      "loss": 0.5322,
      "step": 1260
    },
    {
      "epoch": 0.61,
      "learning_rate": 9.763277169231307e-06,
      "loss": 0.535,
      "step": 1270
    },
    {
      "epoch": 0.62,
      "learning_rate": 9.757123099594078e-06,
      "loss": 0.5216,
      "step": 1280
    },
    {
      "epoch": 0.62,
      "learning_rate": 9.750892049842088e-06,
      "loss": 0.5189,
      "step": 1290
    },
    {
      "epoch": 0.63,
      "learning_rate": 9.744584120806644e-06,
      "loss": 0.5265,
      "step": 1300
    },
    {
      "epoch": 0.63,
      "learning_rate": 9.738199414563113e-06,
      "loss": 0.5274,
      "step": 1310
    },
    {
      "epoch": 0.64,
      "learning_rate": 9.731738034429281e-06,
      "loss": 0.5196,
      "step": 1320
    },
    {
      "epoch": 0.64,
      "learning_rate": 9.725200084963676e-06,
      "loss": 0.5204,
      "step": 1330
    },
    {
      "epoch": 0.64,
      "learning_rate": 9.71858567196387e-06,
      "loss": 0.5323,
      "step": 1340
    },
    {
      "epoch": 0.65,
      "learning_rate": 9.711894902464789e-06,
      "loss": 0.5406,
      "step": 1350
    },
    {
      "epoch": 0.65,
      "learning_rate": 9.705127884736947e-06,
      "loss": 0.5274,
      "step": 1360
    },
    {
      "epoch": 0.66,
      "learning_rate": 9.698284728284723e-06,
      "loss": 0.5152,
      "step": 1370
    },
    {
      "epoch": 0.66,
      "learning_rate": 9.691365543844578e-06,
      "loss": 0.5235,
      "step": 1380
    },
    {
      "epoch": 0.67,
      "learning_rate": 9.684370443383262e-06,
      "loss": 0.5234,
      "step": 1390
    },
    {
      "epoch": 0.67,
      "learning_rate": 9.677299540096002e-06,
      "loss": 0.5128,
      "step": 1400
    },
    {
      "epoch": 0.68,
      "learning_rate": 9.670152948404674e-06,
      "loss": 0.5438,
      "step": 1410
    },
    {
      "epoch": 0.68,
      "learning_rate": 9.662930783955948e-06,
      "loss": 0.5242,
      "step": 1420
    },
    {
      "epoch": 0.69,
      "learning_rate": 9.65563316361942e-06,
      "loss": 0.5387,
      "step": 1430
    },
    {
      "epoch": 0.69,
      "learning_rate": 9.648260205485718e-06,
      "loss": 0.5148,
      "step": 1440
    },
    {
      "epoch": 0.7,
      "learning_rate": 9.640812028864588e-06,
      "loss": 0.51,
      "step": 1450
    },
    {
      "epoch": 0.7,
      "learning_rate": 9.633288754282972e-06,
      "loss": 0.4974,
      "step": 1460
    },
    {
      "epoch": 0.71,
      "learning_rate": 9.625690503483051e-06,
      "loss": 0.5009,
      "step": 1470
    },
    {
      "epoch": 0.71,
      "learning_rate": 9.618017399420277e-06,
      "loss": 0.5155,
      "step": 1480
    },
    {
      "epoch": 0.72,
      "learning_rate": 9.610269566261382e-06,
      "loss": 0.5268,
      "step": 1490
    },
    {
      "epoch": 0.72,
      "learning_rate": 9.602447129382368e-06,
      "loss": 0.5118,
      "step": 1500
    },
    {
      "epoch": 0.72,
      "eval_webgpt_accuracy": 0.5863125638406538,
      "eval_webgpt_loss": 0.6783966422080994,
      "eval_webgpt_runtime": 39.014,
      "eval_webgpt_samples_per_second": 100.374,
      "eval_webgpt_steps_per_second": 1.333,
      "step": 1500
    },
    {
      "epoch": 0.72,
      "eval_hfsummary_accuracy": 0.6642082036091044,
      "eval_hfsummary_loss": 0.6211047172546387,
      "eval_hfsummary_runtime": 981.3434,
      "eval_hfsummary_samples_per_second": 33.712,
      "eval_hfsummary_steps_per_second": 0.444,
      "step": 1500
    },
    {
      "epoch": 0.72,
      "eval_anthropic_rlhf_accuracy": 0.6421889616463985,
      "eval_anthropic_rlhf_loss": 0.6418160796165466,
      "eval_anthropic_rlhf_runtime": 91.7717,
      "eval_anthropic_rlhf_samples_per_second": 93.188,
      "eval_anthropic_rlhf_steps_per_second": 1.231,
      "step": 1500
    },
    {
      "epoch": 0.72,
      "eval_gptsynthetic_accuracy": 0.9969834087481146,
      "eval_gptsynthetic_loss": 0.01542825810611248,
      "eval_gptsynthetic_runtime": 31.9538,
      "eval_gptsynthetic_samples_per_second": 103.743,
      "eval_gptsynthetic_steps_per_second": 1.377,
      "step": 1500
    },
    {
      "epoch": 0.73,
      "learning_rate": 9.594550215366485e-06,
      "loss": 0.517,
      "step": 1510
    },
    {
      "epoch": 0.73,
      "learning_rate": 9.586578952002173e-06,
      "loss": 0.527,
      "step": 1520
    },
    {
      "epoch": 0.74,
      "learning_rate": 9.578533468281001e-06,
      "loss": 0.518,
      "step": 1530
    },
    {
      "epoch": 0.74,
      "learning_rate": 9.570413894395577e-06,
      "loss": 0.5312,
      "step": 1540
    },
    {
      "epoch": 0.75,
      "learning_rate": 9.562220361737437e-06,
      "loss": 0.5052,
      "step": 1550
    },
    {
      "epoch": 0.75,
      "learning_rate": 9.553953002894933e-06,
      "loss": 0.5263,
      "step": 1560
    },
    {
      "epoch": 0.76,
      "learning_rate": 9.545611951651068e-06,
      "loss": 0.5323,
      "step": 1570
    },
    {
      "epoch": 0.76,
      "learning_rate": 9.537197342981346e-06,
      "loss": 0.5091,
      "step": 1580
    },
    {
      "epoch": 0.77,
      "learning_rate": 9.52870931305158e-06,
      "loss": 0.5316,
      "step": 1590
    },
    {
      "epoch": 0.77,
      "learning_rate": 9.520147999215697e-06,
      "loss": 0.5067,
      "step": 1600
    },
    {
      "epoch": 0.77,
      "learning_rate": 9.511513540013498e-06,
      "loss": 0.5102,
      "step": 1610
    },
    {
      "epoch": 0.78,
      "learning_rate": 9.502806075168441e-06,
      "loss": 0.5119,
      "step": 1620
    },
    {
      "epoch": 0.78,
      "learning_rate": 9.494025745585357e-06,
      "loss": 0.5256,
      "step": 1630
    },
    {
      "epoch": 0.79,
      "learning_rate": 9.485172693348186e-06,
      "loss": 0.5119,
      "step": 1640
    },
    {
      "epoch": 0.79,
      "learning_rate": 9.476247061717666e-06,
      "loss": 0.5071,
      "step": 1650
    },
    {
      "epoch": 0.8,
      "learning_rate": 9.467248995129022e-06,
      "loss": 0.5241,
      "step": 1660
    },
    {
      "epoch": 0.8,
      "learning_rate": 9.45817863918963e-06,
      "loss": 0.5155,
      "step": 1670
    },
    {
      "epoch": 0.81,
      "learning_rate": 9.449036140676652e-06,
      "loss": 0.4948,
      "step": 1680
    },
    {
      "epoch": 0.81,
      "learning_rate": 9.43982164753467e-06,
      "loss": 0.5125,
      "step": 1690
    },
    {
      "epoch": 0.82,
      "learning_rate": 9.43053530887329e-06,
      "loss": 0.5095,
      "step": 1700
    },
    {
      "epoch": 0.82,
      "learning_rate": 9.42117727496472e-06,
      "loss": 0.5122,
      "step": 1710
    },
    {
      "epoch": 0.83,
      "learning_rate": 9.411747697241356e-06,
      "loss": 0.4946,
      "step": 1720
    },
    {
      "epoch": 0.83,
      "learning_rate": 9.402246728293313e-06,
      "loss": 0.5244,
      "step": 1730
    },
    {
      "epoch": 0.84,
      "learning_rate": 9.392674521865968e-06,
      "loss": 0.4982,
      "step": 1740
    },
    {
      "epoch": 0.84,
      "learning_rate": 9.383031232857466e-06,
      "loss": 0.52,
      "step": 1750
    },
    {
      "epoch": 0.85,
      "learning_rate": 9.373317017316218e-06,
      "loss": 0.5022,
      "step": 1760
    },
    {
      "epoch": 0.85,
      "learning_rate": 9.363532032438372e-06,
      "loss": 0.5131,
      "step": 1770
    },
    {
      "epoch": 0.86,
      "learning_rate": 9.353676436565271e-06,
      "loss": 0.5072,
      "step": 1780
    },
    {
      "epoch": 0.86,
      "learning_rate": 9.343750389180887e-06,
      "loss": 0.5201,
      "step": 1790
    },
    {
      "epoch": 0.87,
      "learning_rate": 9.333754050909251e-06,
      "loss": 0.4917,
      "step": 1800
    },
    {
      "epoch": 0.87,
      "learning_rate": 9.32368758351184e-06,
      "loss": 0.4938,
      "step": 1810
    },
    {
      "epoch": 0.88,
      "learning_rate": 9.313551149884968e-06,
      "loss": 0.5183,
      "step": 1820
    },
    {
      "epoch": 0.88,
      "learning_rate": 9.303344914057145e-06,
      "loss": 0.4944,
      "step": 1830
    },
    {
      "epoch": 0.89,
      "learning_rate": 9.293069041186426e-06,
      "loss": 0.5088,
      "step": 1840
    },
    {
      "epoch": 0.89,
      "learning_rate": 9.282723697557746e-06,
      "loss": 0.5025,
      "step": 1850
    },
    {
      "epoch": 0.9,
      "learning_rate": 9.272309050580207e-06,
      "loss": 0.5302,
      "step": 1860
    },
    {
      "epoch": 0.9,
      "learning_rate": 9.261825268784395e-06,
      "loss": 0.4862,
      "step": 1870
    },
    {
      "epoch": 0.9,
      "learning_rate": 9.25127252181963e-06,
      "loss": 0.5122,
      "step": 1880
    },
    {
      "epoch": 0.91,
      "learning_rate": 9.240650980451244e-06,
      "loss": 0.512,
      "step": 1890
    },
    {
      "epoch": 0.91,
      "learning_rate": 9.229960816557793e-06,
      "loss": 0.4975,
      "step": 1900
    },
    {
      "epoch": 0.92,
      "learning_rate": 9.219202203128293e-06,
      "loss": 0.4917,
      "step": 1910
    },
    {
      "epoch": 0.92,
      "learning_rate": 9.20837531425942e-06,
      "loss": 0.4892,
      "step": 1920
    },
    {
      "epoch": 0.93,
      "learning_rate": 9.197480325152677e-06,
      "loss": 0.495,
      "step": 1930
    },
    {
      "epoch": 0.93,
      "learning_rate": 9.18651741211158e-06,
      "loss": 0.5009,
      "step": 1940
    },
    {
      "epoch": 0.94,
      "learning_rate": 9.175486752538792e-06,
      "loss": 0.526,
      "step": 1950
    },
    {
      "epoch": 0.94,
      "learning_rate": 9.164388524933253e-06,
      "loss": 0.505,
      "step": 1960
    },
    {
      "epoch": 0.95,
      "learning_rate": 9.1532229088873e-06,
      "loss": 0.4911,
      "step": 1970
    },
    {
      "epoch": 0.95,
      "learning_rate": 9.141990085083746e-06,
      "loss": 0.4796,
      "step": 1980
    },
    {
      "epoch": 0.96,
      "learning_rate": 9.130690235292969e-06,
      "loss": 0.5006,
      "step": 1990
    },
    {
      "epoch": 0.96,
      "learning_rate": 9.119323542369969e-06,
      "loss": 0.4968,
      "step": 2000
    },
    {
      "epoch": 0.96,
      "eval_webgpt_accuracy": 0.5903983656792645,
      "eval_webgpt_loss": 0.6728653311729431,
      "eval_webgpt_runtime": 39.1014,
      "eval_webgpt_samples_per_second": 100.15,
      "eval_webgpt_steps_per_second": 1.33,
      "step": 2000
    },
    {
      "epoch": 0.96,
      "eval_hfsummary_accuracy": 0.6710999607048937,
      "eval_hfsummary_loss": 0.646239697933197,
      "eval_hfsummary_runtime": 984.6253,
      "eval_hfsummary_samples_per_second": 33.6,
      "eval_hfsummary_steps_per_second": 0.443,
      "step": 2000
    },
    {
      "epoch": 0.96,
      "eval_anthropic_rlhf_accuracy": 0.6413704396632367,
      "eval_anthropic_rlhf_loss": 0.6394137740135193,
      "eval_anthropic_rlhf_runtime": 92.1998,
      "eval_anthropic_rlhf_samples_per_second": 92.755,
      "eval_anthropic_rlhf_steps_per_second": 1.226,
      "step": 2000
    },
    {
      "epoch": 0.96,
      "eval_gptsynthetic_accuracy": 0.9972850678733032,
      "eval_gptsynthetic_loss": 0.022552527487277985,
      "eval_gptsynthetic_runtime": 32.0814,
      "eval_gptsynthetic_samples_per_second": 103.331,
      "eval_gptsynthetic_steps_per_second": 1.372,
      "step": 2000
    },
    {
      "epoch": 0.97,
      "learning_rate": 9.107890190251403e-06,
      "loss": 0.4929,
      "step": 2010
    },
    {
      "epoch": 0.97,
      "learning_rate": 9.096390363952614e-06,
      "loss": 0.498,
      "step": 2020
    },
    {
      "epoch": 0.98,
      "learning_rate": 9.084824249564635e-06,
      "loss": 0.5041,
      "step": 2030
    },
    {
      "epoch": 0.98,
      "learning_rate": 9.073192034251174e-06,
      "loss": 0.4995,
      "step": 2040
    },
    {
      "epoch": 0.99,
      "learning_rate": 9.061493906245596e-06,
      "loss": 0.5,
      "step": 2050
    },
    {
      "epoch": 0.99,
      "learning_rate": 9.049730054847865e-06,
      "loss": 0.4851,
      "step": 2060
    },
    {
      "epoch": 1.0,
      "learning_rate": 9.037900670421484e-06,
      "loss": 0.5055,
      "step": 2070
    },
    {
      "epoch": 1.0,
      "learning_rate": 9.02600594439042e-06,
      "loss": 0.5374,
      "step": 2080
    },
    {
      "epoch": 1.01,
      "learning_rate": 9.014046069236005e-06,
      "loss": 0.4931,
      "step": 2090
    },
    {
      "epoch": 1.01,
      "learning_rate": 9.00202123849381e-06,
      "loss": 0.4837,
      "step": 2100
    },
    {
      "epoch": 1.02,
      "learning_rate": 8.989931646750533e-06,
      "loss": 0.4903,
      "step": 2110
    },
    {
      "epoch": 1.02,
      "learning_rate": 8.97777748964083e-06,
      "loss": 0.4727,
      "step": 2120
    },
    {
      "epoch": 1.03,
      "learning_rate": 8.965558963844165e-06,
      "loss": 0.479,
      "step": 2130
    },
    {
      "epoch": 1.03,
      "learning_rate": 8.953276267081618e-06,
      "loss": 0.4769,
      "step": 2140
    },
    {
      "epoch": 1.04,
      "learning_rate": 8.940929598112687e-06,
      "loss": 0.4967,
      "step": 2150
    },
    {
      "epoch": 1.04,
      "learning_rate": 8.928519156732074e-06,
      "loss": 0.4734,
      "step": 2160
    },
    {
      "epoch": 1.04,
      "learning_rate": 8.916045143766455e-06,
      "loss": 0.5009,
      "step": 2170
    },
    {
      "epoch": 1.05,
      "learning_rate": 8.90350776107122e-06,
      "loss": 0.469,
      "step": 2180
    },
    {
      "epoch": 1.05,
      "learning_rate": 8.890907211527213e-06,
      "loss": 0.4942,
      "step": 2190
    },
    {
      "epoch": 1.06,
      "learning_rate": 8.878243699037452e-06,
      "loss": 0.468,
      "step": 2200
    },
    {
      "epoch": 1.06,
      "learning_rate": 8.865517428523826e-06,
      "loss": 0.4829,
      "step": 2210
    },
    {
      "epoch": 1.07,
      "learning_rate": 8.852728605923774e-06,
      "loss": 0.501,
      "step": 2220
    },
    {
      "epoch": 1.07,
      "learning_rate": 8.839877438186966e-06,
      "loss": 0.4659,
      "step": 2230
    },
    {
      "epoch": 1.08,
      "learning_rate": 8.826964133271937e-06,
      "loss": 0.4634,
      "step": 2240
    },
    {
      "epoch": 1.08,
      "learning_rate": 8.813988900142734e-06,
      "loss": 0.4656,
      "step": 2250
    },
    {
      "epoch": 1.09,
      "learning_rate": 8.800951948765531e-06,
      "loss": 0.4709,
      "step": 2260
    },
    {
      "epoch": 1.09,
      "learning_rate": 8.787853490105233e-06,
      "loss": 0.4657,
      "step": 2270
    },
    {
      "epoch": 1.1,
      "learning_rate": 8.774693736122055e-06,
      "loss": 0.4773,
      "step": 2280
    },
    {
      "epoch": 1.1,
      "learning_rate": 8.761472899768101e-06,
      "loss": 0.4796,
      "step": 2290
    },
    {
      "epoch": 1.11,
      "learning_rate": 8.748191194983919e-06,
      "loss": 0.4823,
      "step": 2300
    },
    {
      "epoch": 1.11,
      "learning_rate": 8.734848836695023e-06,
      "loss": 0.4655,
      "step": 2310
    },
    {
      "epoch": 1.12,
      "learning_rate": 8.721446040808438e-06,
      "loss": 0.4838,
      "step": 2320
    },
    {
      "epoch": 1.12,
      "learning_rate": 8.707983024209186e-06,
      "loss": 0.4927,
      "step": 2330
    },
    {
      "epoch": 1.13,
      "learning_rate": 8.694460004756791e-06,
      "loss": 0.4772,
      "step": 2340
    },
    {
      "epoch": 1.13,
      "learning_rate": 8.680877201281746e-06,
      "loss": 0.4792,
      "step": 2350
    },
    {
      "epoch": 1.14,
      "learning_rate": 8.66723483358197e-06,
      "loss": 0.4739,
      "step": 2360
    },
    {
      "epoch": 1.14,
      "learning_rate": 8.653533122419257e-06,
      "loss": 0.4876,
      "step": 2370
    },
    {
      "epoch": 1.15,
      "learning_rate": 8.639772289515705e-06,
      "loss": 0.4471,
      "step": 2380
    },
    {
      "epoch": 1.15,
      "learning_rate": 8.625952557550119e-06,
      "loss": 0.4641,
      "step": 2390
    },
    {
      "epoch": 1.16,
      "learning_rate": 8.612074150154412e-06,
      "loss": 0.468,
      "step": 2400
    },
    {
      "epoch": 1.16,
      "learning_rate": 8.598137291909996e-06,
      "loss": 0.4665,
      "step": 2410
    },
    {
      "epoch": 1.17,
      "learning_rate": 8.584142208344128e-06,
      "loss": 0.452,
      "step": 2420
    },
    {
      "epoch": 1.17,
      "learning_rate": 8.570089125926279e-06,
      "loss": 0.4811,
      "step": 2430
    },
    {
      "epoch": 1.17,
      "learning_rate": 8.555978272064456e-06,
      "loss": 0.4823,
      "step": 2440
    },
    {
      "epoch": 1.18,
      "learning_rate": 8.541809875101535e-06,
      "loss": 0.4692,
      "step": 2450
    },
    {
      "epoch": 1.18,
      "learning_rate": 8.527584164311548e-06,
      "loss": 0.4704,
      "step": 2460
    },
    {
      "epoch": 1.19,
      "learning_rate": 8.513301369895998e-06,
      "loss": 0.4721,
      "step": 2470
    },
    {
      "epoch": 1.19,
      "learning_rate": 8.498961722980106e-06,
      "loss": 0.4416,
      "step": 2480
    },
    {
      "epoch": 1.2,
      "learning_rate": 8.48456545560909e-06,
      "loss": 0.4708,
      "step": 2490
    },
    {
      "epoch": 1.2,
      "learning_rate": 8.470112800744406e-06,
      "loss": 0.4855,
      "step": 2500
    },
    {
      "epoch": 1.2,
      "eval_webgpt_accuracy": 0.5863125638406538,
      "eval_webgpt_loss": 0.6755186319351196,
      "eval_webgpt_runtime": 39.3033,
      "eval_webgpt_samples_per_second": 99.635,
      "eval_webgpt_steps_per_second": 1.323,
      "step": 2500
    },
    {
      "epoch": 1.2,
      "eval_hfsummary_accuracy": 0.6743342502191458,
      "eval_hfsummary_loss": 0.6676438450813293,
      "eval_hfsummary_runtime": 986.7916,
      "eval_hfsummary_samples_per_second": 33.526,
      "eval_hfsummary_steps_per_second": 0.442,
      "step": 2500
    },
    {
      "epoch": 1.2,
      "eval_anthropic_rlhf_accuracy": 0.6465154349859682,
      "eval_anthropic_rlhf_loss": 0.635563850402832,
      "eval_anthropic_rlhf_runtime": 92.2813,
      "eval_anthropic_rlhf_samples_per_second": 92.673,
      "eval_anthropic_rlhf_steps_per_second": 1.225,
      "step": 2500
    },
    {
      "epoch": 1.2,
      "eval_gptsynthetic_accuracy": 0.9975867269984917,
      "eval_gptsynthetic_loss": 0.016244370490312576,
      "eval_gptsynthetic_runtime": 32.0562,
      "eval_gptsynthetic_samples_per_second": 103.412,
      "eval_gptsynthetic_steps_per_second": 1.373,
      "step": 2500
    },
    {
      "epoch": 1.21,
      "learning_rate": 8.455603992259972e-06,
      "loss": 0.4741,
      "step": 2510
    },
    {
      "epoch": 1.21,
      "learning_rate": 8.441039264938393e-06,
      "loss": 0.4763,
      "step": 2520
    },
    {
      "epoch": 1.22,
      "learning_rate": 8.426418854467154e-06,
      "loss": 0.4687,
      "step": 2530
    },
    {
      "epoch": 1.22,
      "learning_rate": 8.411742997434809e-06,
      "loss": 0.4707,
      "step": 2540
    },
    {
      "epoch": 1.23,
      "learning_rate": 8.39701193132715e-06,
      "loss": 0.4658,
      "step": 2550
    },
    {
      "epoch": 1.23,
      "learning_rate": 8.382225894523373e-06,
      "loss": 0.4709,
      "step": 2560
    },
    {
      "epoch": 1.24,
      "learning_rate": 8.367385126292207e-06,
      "loss": 0.475,
      "step": 2570
    },
    {
      "epoch": 1.24,
      "learning_rate": 8.352489866788051e-06,
      "loss": 0.4764,
      "step": 2580
    },
    {
      "epoch": 1.25,
      "learning_rate": 8.337540357047087e-06,
      "loss": 0.4954,
      "step": 2590
    },
    {
      "epoch": 1.25,
      "learning_rate": 8.32253683898338e-06,
      "loss": 0.4626,
      "step": 2600
    },
    {
      "epoch": 1.26,
      "learning_rate": 8.307479555384954e-06,
      "loss": 0.4724,
      "step": 2610
    },
    {
      "epoch": 1.26,
      "learning_rate": 8.29236874990988e-06,
      "loss": 0.4564,
      "step": 2620
    },
    {
      "epoch": 1.27,
      "learning_rate": 8.27720466708232e-06,
      "loss": 0.4582,
      "step": 2630
    },
    {
      "epoch": 1.27,
      "learning_rate": 8.261987552288573e-06,
      "loss": 0.482,
      "step": 2640
    },
    {
      "epoch": 1.28,
      "learning_rate": 8.246717651773108e-06,
      "loss": 0.4907,
      "step": 2650
    },
    {
      "epoch": 1.28,
      "learning_rate": 8.231395212634572e-06,
      "loss": 0.4576,
      "step": 2660
    },
    {
      "epoch": 1.29,
      "learning_rate": 8.216020482821797e-06,
      "loss": 0.4544,
      "step": 2670
    },
    {
      "epoch": 1.29,
      "learning_rate": 8.200593711129796e-06,
      "loss": 0.4665,
      "step": 2680
    },
    {
      "epoch": 1.3,
      "learning_rate": 8.18511514719571e-06,
      "loss": 0.4729,
      "step": 2690
    },
    {
      "epoch": 1.3,
      "learning_rate": 8.1695850414948e-06,
      "loss": 0.4891,
      "step": 2700
    },
    {
      "epoch": 1.3,
      "learning_rate": 8.154003645336378e-06,
      "loss": 0.4538,
      "step": 2710
    },
    {
      "epoch": 1.31,
      "learning_rate": 8.138371210859732e-06,
      "loss": 0.476,
      "step": 2720
    },
    {
      "epoch": 1.31,
      "learning_rate": 8.122687991030067e-06,
      "loss": 0.4519,
      "step": 2730
    },
    {
      "epoch": 1.32,
      "learning_rate": 8.106954239634394e-06,
      "loss": 0.4644,
      "step": 2740
    },
    {
      "epoch": 1.32,
      "learning_rate": 8.091170211277437e-06,
      "loss": 0.4551,
      "step": 2750
    },
    {
      "epoch": 1.33,
      "learning_rate": 8.075336161377492e-06,
      "loss": 0.4765,
      "step": 2760
    },
    {
      "epoch": 1.33,
      "learning_rate": 8.059452346162321e-06,
      "loss": 0.4481,
      "step": 2770
    },
    {
      "epoch": 1.34,
      "learning_rate": 8.043519022664984e-06,
      "loss": 0.47,
      "step": 2780
    },
    {
      "epoch": 1.34,
      "learning_rate": 8.027536448719692e-06,
      "loss": 0.4955,
      "step": 2790
    },
    {
      "epoch": 1.35,
      "learning_rate": 8.011504882957626e-06,
      "loss": 0.4688,
      "step": 2800
    },
    {
      "epoch": 1.35,
      "learning_rate": 7.99542458480276e-06,
      "loss": 0.447,
      "step": 2810
    },
    {
      "epoch": 1.36,
      "learning_rate": 7.979295814467656e-06,
      "loss": 0.4581,
      "step": 2820
    },
    {
      "epoch": 1.36,
      "learning_rate": 7.963118832949259e-06,
      "loss": 0.4857,
      "step": 2830
    },
    {
      "epoch": 1.37,
      "learning_rate": 7.94689390202467e-06,
      "loss": 0.458,
      "step": 2840
    },
    {
      "epoch": 1.37,
      "learning_rate": 7.930621284246914e-06,
      "loss": 0.4797,
      "step": 2850
    },
    {
      "epoch": 1.38,
      "learning_rate": 7.914301242940678e-06,
      "loss": 0.4722,
      "step": 2860
    },
    {
      "epoch": 1.38,
      "learning_rate": 7.897934042198073e-06,
      "loss": 0.4499,
      "step": 2870
    },
    {
      "epoch": 1.39,
      "learning_rate": 7.881519946874336e-06,
      "loss": 0.4457,
      "step": 2880
    },
    {
      "epoch": 1.39,
      "learning_rate": 7.865059222583565e-06,
      "loss": 0.4563,
      "step": 2890
    },
    {
      "epoch": 1.4,
      "learning_rate": 7.848552135694407e-06,
      "loss": 0.4753,
      "step": 2900
    },
    {
      "epoch": 1.4,
      "learning_rate": 7.83199895332575e-06,
      "loss": 0.4595,
      "step": 2910
    },
    {
      "epoch": 1.41,
      "learning_rate": 7.815399943342408e-06,
      "loss": 0.4529,
      "step": 2920
    },
    {
      "epoch": 1.41,
      "learning_rate": 7.798755374350774e-06,
      "loss": 0.4634,
      "step": 2930
    },
    {
      "epoch": 1.42,
      "learning_rate": 7.782065515694486e-06,
      "loss": 0.4671,
      "step": 2940
    },
    {
      "epoch": 1.42,
      "learning_rate": 7.765330637450057e-06,
      "loss": 0.4921,
      "step": 2950
    },
    {
      "epoch": 1.42,
      "learning_rate": 7.748551010422518e-06,
      "loss": 0.4841,
      "step": 2960
    },
    {
      "epoch": 1.43,
      "learning_rate": 7.731726906141024e-06,
      "loss": 0.4348,
      "step": 2970
    },
    {
      "epoch": 1.43,
      "learning_rate": 7.71485859685446e-06,
      "loss": 0.4612,
      "step": 2980
    },
    {
      "epoch": 1.44,
      "learning_rate": 7.697946355527044e-06,
      "loss": 0.4517,
      "step": 2990
    },
    {
      "epoch": 1.44,
      "learning_rate": 7.680990455833907e-06,
      "loss": 0.4514,
      "step": 3000
    },
    {
      "epoch": 1.44,
      "eval_webgpt_accuracy": 0.5827374872318692,
      "eval_webgpt_loss": 0.6752701997756958,
      "eval_webgpt_runtime": 39.1231,
      "eval_webgpt_samples_per_second": 100.094,
      "eval_webgpt_steps_per_second": 1.329,
      "step": 3000
    },
    {
      "epoch": 1.44,
      "eval_hfsummary_accuracy": 0.6794123870265696,
      "eval_hfsummary_loss": 0.6801344156265259,
      "eval_hfsummary_runtime": 986.8465,
      "eval_hfsummary_samples_per_second": 33.524,
      "eval_hfsummary_steps_per_second": 0.442,
      "step": 3000
    },
    {
      "epoch": 1.44,
      "eval_anthropic_rlhf_accuracy": 0.657156220767072,
      "eval_anthropic_rlhf_loss": 0.6261478662490845,
      "eval_anthropic_rlhf_runtime": 92.2987,
      "eval_anthropic_rlhf_samples_per_second": 92.656,
      "eval_anthropic_rlhf_steps_per_second": 1.224,
      "step": 3000
    },
    {
      "epoch": 1.44,
      "eval_gptsynthetic_accuracy": 0.9969834087481146,
      "eval_gptsynthetic_loss": 0.009839520789682865,
      "eval_gptsynthetic_runtime": 32.2002,
      "eval_gptsynthetic_samples_per_second": 102.95,
      "eval_gptsynthetic_steps_per_second": 1.366,
      "step": 3000
    },
    {
      "epoch": 1.45,
      "learning_rate": 7.663991172156662e-06,
      "loss": 0.4668,
      "step": 3010
    },
    {
      "epoch": 1.45,
      "learning_rate": 7.646948779578961e-06,
      "loss": 0.4689,
      "step": 3020
    },
    {
      "epoch": 1.46,
      "learning_rate": 7.62986355388205e-06,
      "loss": 0.4544,
      "step": 3030
    },
    {
      "epoch": 1.46,
      "learning_rate": 7.612735771540308e-06,
      "loss": 0.4549,
      "step": 3040
    },
    {
      "epoch": 1.47,
      "learning_rate": 7.59556570971676e-06,
      "loss": 0.4738,
      "step": 3050
    },
    {
      "epoch": 1.47,
      "learning_rate": 7.5783536462586076e-06,
      "loss": 0.4585,
      "step": 3060
    },
    {
      "epoch": 1.48,
      "learning_rate": 7.561099859692722e-06,
      "loss": 0.4782,
      "step": 3070
    },
    {
      "epoch": 1.48,
      "learning_rate": 7.543804629221143e-06,
      "loss": 0.475,
      "step": 3080
    },
    {
      "epoch": 1.49,
      "learning_rate": 7.526468234716557e-06,
      "loss": 0.465,
      "step": 3090
    },
    {
      "epoch": 1.49,
      "learning_rate": 7.509090956717775e-06,
      "loss": 0.4689,
      "step": 3100
    },
    {
      "epoch": 1.5,
      "learning_rate": 7.49167307642518e-06,
      "loss": 0.4652,
      "step": 3110
    },
    {
      "epoch": 1.5,
      "learning_rate": 7.474214875696188e-06,
      "loss": 0.4409,
      "step": 3120
    },
    {
      "epoch": 1.51,
      "learning_rate": 7.456716637040683e-06,
      "loss": 0.4637,
      "step": 3130
    },
    {
      "epoch": 1.51,
      "learning_rate": 7.439178643616447e-06,
      "loss": 0.4402,
      "step": 3140
    },
    {
      "epoch": 1.52,
      "learning_rate": 7.421601179224574e-06,
      "loss": 0.4522,
      "step": 3150
    },
    {
      "epoch": 1.52,
      "learning_rate": 7.403984528304884e-06,
      "loss": 0.4457,
      "step": 3160
    },
    {
      "epoch": 1.53,
      "learning_rate": 7.386328975931311e-06,
      "loss": 0.4591,
      "step": 3170
    },
    {
      "epoch": 1.53,
      "learning_rate": 7.368634807807298e-06,
      "loss": 0.4509,
      "step": 3180
    },
    {
      "epoch": 1.54,
      "learning_rate": 7.3509023102611745e-06,
      "loss": 0.4404,
      "step": 3190
    },
    {
      "epoch": 1.54,
      "learning_rate": 7.33313177024151e-06,
      "loss": 0.461,
      "step": 3200
    },
    {
      "epoch": 1.55,
      "learning_rate": 7.315323475312489e-06,
      "loss": 0.4479,
      "step": 3210
    },
    {
      "epoch": 1.55,
      "learning_rate": 7.297477713649247e-06,
      "loss": 0.4604,
      "step": 3220
    },
    {
      "epoch": 1.55,
      "learning_rate": 7.279594774033205e-06,
      "loss": 0.4488,
      "step": 3230
    },
    {
      "epoch": 1.56,
      "learning_rate": 7.261674945847406e-06,
      "loss": 0.4634,
      "step": 3240
    },
    {
      "epoch": 1.56,
      "learning_rate": 7.2437185190718205e-06,
      "loss": 0.4503,
      "step": 3250
    },
    {
      "epoch": 1.57,
      "learning_rate": 7.225725784278663e-06,
      "loss": 0.4605,
      "step": 3260
    },
    {
      "epoch": 1.57,
      "learning_rate": 7.207697032627689e-06,
      "loss": 0.446,
      "step": 3270
    },
    {
      "epoch": 1.58,
      "learning_rate": 7.18963255586148e-06,
      "loss": 0.4385,
      "step": 3280
    },
    {
      "epoch": 1.58,
      "learning_rate": 7.17153264630072e-06,
      "loss": 0.45,
      "step": 3290
    },
    {
      "epoch": 1.59,
      "learning_rate": 7.153397596839475e-06,
      "loss": 0.4611,
      "step": 3300
    },
    {
      "epoch": 1.59,
      "learning_rate": 7.1352277009404454e-06,
      "loss": 0.4495,
      "step": 3310
    },
    {
      "epoch": 1.6,
      "learning_rate": 7.117023252630216e-06,
      "loss": 0.4377,
      "step": 3320
    },
    {
      "epoch": 1.6,
      "learning_rate": 7.098784546494507e-06,
      "loss": 0.4517,
      "step": 3330
    },
    {
      "epoch": 1.61,
      "learning_rate": 7.080511877673397e-06,
      "loss": 0.4461,
      "step": 3340
    },
    {
      "epoch": 1.61,
      "learning_rate": 7.062205541856552e-06,
      "loss": 0.4587,
      "step": 3350
    },
    {
      "epoch": 1.62,
      "learning_rate": 7.04386583527844e-06,
      "loss": 0.4473,
      "step": 3360
    },
    {
      "epoch": 1.62,
      "learning_rate": 7.025493054713542e-06,
      "loss": 0.4516,
      "step": 3370
    },
    {
      "epoch": 1.63,
      "learning_rate": 7.007087497471532e-06,
      "loss": 0.4698,
      "step": 3380
    },
    {
      "epoch": 1.63,
      "learning_rate": 6.988649461392491e-06,
      "loss": 0.4415,
      "step": 3390
    },
    {
      "epoch": 1.64,
      "learning_rate": 6.970179244842074e-06,
      "loss": 0.4593,
      "step": 3400
    },
    {
      "epoch": 1.64,
      "learning_rate": 6.9516771467066726e-06,
      "loss": 0.4428,
      "step": 3410
    },
    {
      "epoch": 1.65,
      "learning_rate": 6.933143466388599e-06,
      "loss": 0.4606,
      "step": 3420
    },
    {
      "epoch": 1.65,
      "learning_rate": 6.914578503801228e-06,
      "loss": 0.457,
      "step": 3430
    },
    {
      "epoch": 1.66,
      "learning_rate": 6.8959825593641425e-06,
      "loss": 0.4408,
      "step": 3440
    },
    {
      "epoch": 1.66,
      "learning_rate": 6.87735593399828e-06,
      "loss": 0.4636,
      "step": 3450
    },
    {
      "epoch": 1.67,
      "learning_rate": 6.858698929121058e-06,
      "loss": 0.4519,
      "step": 3460
    },
    {
      "epoch": 1.67,
      "learning_rate": 6.840011846641497e-06,
      "loss": 0.4586,
      "step": 3470
    },
    {
      "epoch": 1.68,
      "learning_rate": 6.821294988955334e-06,
      "loss": 0.4316,
      "step": 3480
    },
    {
      "epoch": 1.68,
      "learning_rate": 6.802548658940136e-06,
      "loss": 0.4685,
      "step": 3490
    },
    {
      "epoch": 1.68,
      "learning_rate": 6.783773159950386e-06,
      "loss": 0.4427,
      "step": 3500
    },
    {
      "epoch": 1.68,
      "eval_webgpt_accuracy": 0.5957609805924413,
      "eval_webgpt_loss": 0.674083948135376,
      "eval_webgpt_runtime": 39.2401,
      "eval_webgpt_samples_per_second": 99.796,
      "eval_webgpt_steps_per_second": 1.325,
      "step": 3500
    },
    {
      "epoch": 1.68,
      "eval_hfsummary_accuracy": 0.6743040232143397,
      "eval_hfsummary_loss": 0.6820210814476013,
      "eval_hfsummary_runtime": 984.942,
      "eval_hfsummary_samples_per_second": 33.589,
      "eval_hfsummary_steps_per_second": 0.443,
      "step": 3500
    },
    {
      "epoch": 1.68,
      "eval_anthropic_rlhf_accuracy": 0.6555191768007483,
      "eval_anthropic_rlhf_loss": 0.6293210983276367,
      "eval_anthropic_rlhf_runtime": 92.3544,
      "eval_anthropic_rlhf_samples_per_second": 92.6,
      "eval_anthropic_rlhf_steps_per_second": 1.224,
      "step": 3500
    },
    {
      "epoch": 1.68,
      "eval_gptsynthetic_accuracy": 0.9975867269984917,
      "eval_gptsynthetic_loss": 0.01304171234369278,
      "eval_gptsynthetic_runtime": 32.2744,
      "eval_gptsynthetic_samples_per_second": 102.713,
      "eval_gptsynthetic_steps_per_second": 1.363,
      "step": 3500
    },
    {
      "epoch": 1.69,
      "learning_rate": 6.764968795812587e-06,
      "loss": 0.4489,
      "step": 3510
    },
    {
      "epoch": 1.69,
      "learning_rate": 6.746135870820338e-06,
      "loss": 0.4492,
      "step": 3520
    },
    {
      "epoch": 1.7,
      "learning_rate": 6.727274689729411e-06,
      "loss": 0.4482,
      "step": 3530
    },
    {
      "epoch": 1.7,
      "learning_rate": 6.7083855577528234e-06,
      "loss": 0.455,
      "step": 3540
    },
    {
      "epoch": 1.71,
      "learning_rate": 6.689468780555892e-06,
      "loss": 0.4285,
      "step": 3550
    },
    {
      "epoch": 1.71,
      "learning_rate": 6.670524664251291e-06,
      "loss": 0.4485,
      "step": 3560
    },
    {
      "epoch": 1.72,
      "learning_rate": 6.651553515394102e-06,
      "loss": 0.4534,
      "step": 3570
    },
    {
      "epoch": 1.72,
      "learning_rate": 6.632555640976846e-06,
      "loss": 0.4551,
      "step": 3580
    },
    {
      "epoch": 1.73,
      "learning_rate": 6.613531348424518e-06,
      "loss": 0.4409,
      "step": 3590
    },
    {
      "epoch": 1.73,
      "learning_rate": 6.5944809455896145e-06,
      "loss": 0.444,
      "step": 3600
    },
    {
      "epoch": 1.74,
      "learning_rate": 6.575404740747155e-06,
      "loss": 0.4487,
      "step": 3610
    },
    {
      "epoch": 1.74,
      "learning_rate": 6.55630304258968e-06,
      "loss": 0.4339,
      "step": 3620
    },
    {
      "epoch": 1.75,
      "learning_rate": 6.537176160222271e-06,
      "loss": 0.4464,
      "step": 3630
    },
    {
      "epoch": 1.75,
      "learning_rate": 6.518024403157544e-06,
      "loss": 0.4342,
      "step": 3640
    },
    {
      "epoch": 1.76,
      "learning_rate": 6.4988480813106305e-06,
      "loss": 0.4467,
      "step": 3650
    },
    {
      "epoch": 1.76,
      "learning_rate": 6.47964750499418e-06,
      "loss": 0.4536,
      "step": 3660
    },
    {
      "epoch": 1.77,
      "learning_rate": 6.460422984913326e-06,
      "loss": 0.455,
      "step": 3670
    },
    {
      "epoch": 1.77,
      "learning_rate": 6.44117483216066e-06,
      "loss": 0.4372,
      "step": 3680
    },
    {
      "epoch": 1.78,
      "learning_rate": 6.4219033582112026e-06,
      "loss": 0.431,
      "step": 3690
    },
    {
      "epoch": 1.78,
      "learning_rate": 6.402608874917358e-06,
      "loss": 0.438,
      "step": 3700
    },
    {
      "epoch": 1.79,
      "learning_rate": 6.383291694503868e-06,
      "loss": 0.4606,
      "step": 3710
    },
    {
      "epoch": 1.79,
      "learning_rate": 6.3639521295627635e-06,
      "loss": 0.4256,
      "step": 3720
    },
    {
      "epoch": 1.8,
      "learning_rate": 6.344590493048304e-06,
      "loss": 0.4561,
      "step": 3730
    },
    {
      "epoch": 1.8,
      "learning_rate": 6.325207098271907e-06,
      "loss": 0.4479,
      "step": 3740
    },
    {
      "epoch": 1.81,
      "learning_rate": 6.30580225889709e-06,
      "loss": 0.4541,
      "step": 3750
    },
    {
      "epoch": 1.81,
      "learning_rate": 6.286376288934388e-06,
      "loss": 0.4371,
      "step": 3760
    },
    {
      "epoch": 1.81,
      "learning_rate": 6.266929502736268e-06,
      "loss": 0.4312,
      "step": 3770
    },
    {
      "epoch": 1.82,
      "learning_rate": 6.24746221499205e-06,
      "loss": 0.4311,
      "step": 3780
    },
    {
      "epoch": 1.82,
      "learning_rate": 6.227974740722815e-06,
      "loss": 0.4313,
      "step": 3790
    },
    {
      "epoch": 1.83,
      "learning_rate": 6.208467395276298e-06,
      "loss": 0.4704,
      "step": 3800
    },
    {
      "epoch": 1.83,
      "learning_rate": 6.188940494321794e-06,
      "loss": 0.4417,
      "step": 3810
    },
    {
      "epoch": 1.84,
      "learning_rate": 6.169394353845045e-06,
      "loss": 0.4341,
      "step": 3820
    },
    {
      "epoch": 1.84,
      "learning_rate": 6.149829290143129e-06,
      "loss": 0.4558,
      "step": 3830
    },
    {
      "epoch": 1.85,
      "learning_rate": 6.1302456198193415e-06,
      "loss": 0.4511,
      "step": 3840
    },
    {
      "epoch": 1.85,
      "learning_rate": 6.110643659778073e-06,
      "loss": 0.4284,
      "step": 3850
    },
    {
      "epoch": 1.86,
      "learning_rate": 6.091023727219675e-06,
      "loss": 0.4459,
      "step": 3860
    },
    {
      "epoch": 1.86,
      "learning_rate": 6.0713861396353355e-06,
      "loss": 0.4321,
      "step": 3870
    },
    {
      "epoch": 1.87,
      "learning_rate": 6.051731214801939e-06,
      "loss": 0.447,
      "step": 3880
    },
    {
      "epoch": 1.87,
      "learning_rate": 6.032059270776917e-06,
      "loss": 0.4342,
      "step": 3890
    },
    {
      "epoch": 1.88,
      "learning_rate": 6.012370625893112e-06,
      "loss": 0.4391,
      "step": 3900
    },
    {
      "epoch": 1.88,
      "learning_rate": 5.992665598753619e-06,
      "loss": 0.4408,
      "step": 3910
    },
    {
      "epoch": 1.89,
      "learning_rate": 5.972944508226629e-06,
      "loss": 0.4514,
      "step": 3920
    },
    {
      "epoch": 1.89,
      "learning_rate": 5.953207673440278e-06,
      "loss": 0.4748,
      "step": 3930
    },
    {
      "epoch": 1.9,
      "learning_rate": 5.933455413777473e-06,
      "loss": 0.4508,
      "step": 3940
    },
    {
      "epoch": 1.9,
      "learning_rate": 5.913688048870725e-06,
      "loss": 0.4583,
      "step": 3950
    },
    {
      "epoch": 1.91,
      "learning_rate": 5.893905898596986e-06,
      "loss": 0.4331,
      "step": 3960
    },
    {
      "epoch": 1.91,
      "learning_rate": 5.87410928307246e-06,
      "loss": 0.451,
      "step": 3970
    },
    {
      "epoch": 1.92,
      "learning_rate": 5.85429852264743e-06,
      "loss": 0.4601,
      "step": 3980
    },
    {
      "epoch": 1.92,
      "learning_rate": 5.834473937901075e-06,
      "loss": 0.4388,
      "step": 3990
    },
    {
      "epoch": 1.93,
      "learning_rate": 5.814635849636281e-06,
      "loss": 0.4362,
      "step": 4000
    },
    {
      "epoch": 1.93,
      "eval_webgpt_accuracy": 0.5914198161389173,
      "eval_webgpt_loss": 0.672228991985321,
      "eval_webgpt_runtime": 39.1644,
      "eval_webgpt_samples_per_second": 99.989,
      "eval_webgpt_steps_per_second": 1.328,
      "step": 4000
    },
    {
      "epoch": 1.93,
      "eval_hfsummary_accuracy": 0.676631502584409,
      "eval_hfsummary_loss": 0.7084701657295227,
      "eval_hfsummary_runtime": 988.7214,
      "eval_hfsummary_samples_per_second": 33.46,
      "eval_hfsummary_steps_per_second": 0.441,
      "step": 4000
    },
    {
      "epoch": 1.93,
      "eval_anthropic_rlhf_accuracy": 0.6534144059869037,
      "eval_anthropic_rlhf_loss": 0.6274996995925903,
      "eval_anthropic_rlhf_runtime": 92.1951,
      "eval_anthropic_rlhf_samples_per_second": 92.76,
      "eval_anthropic_rlhf_steps_per_second": 1.226,
      "step": 4000
    },
    {
      "epoch": 1.93,
      "eval_gptsynthetic_accuracy": 0.9969834087481146,
      "eval_gptsynthetic_loss": 0.012532506138086319,
      "eval_gptsynthetic_runtime": 32.2302,
      "eval_gptsynthetic_samples_per_second": 102.854,
      "eval_gptsynthetic_steps_per_second": 1.365,
      "step": 4000
    }
  ],
  "max_steps": 4154,
  "num_train_epochs": 2,
  "total_flos": 4.526360616326298e+17,
  "trial_name": null,
  "trial_params": null
}