{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.4998951562172363,
  "eval_steps": 10000,
  "global_step": 1043,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0,
      "learning_rate": 9.904122722914669e-05,
      "loss": 2.4783,
      "step": 10
    },
    {
      "epoch": 0.01,
      "learning_rate": 9.80824544582934e-05,
      "loss": 0.9534,
      "step": 20
    },
    {
      "epoch": 0.01,
      "learning_rate": 9.712368168744008e-05,
      "loss": 0.6314,
      "step": 30
    },
    {
      "epoch": 0.02,
      "learning_rate": 9.616490891658677e-05,
      "loss": 0.5312,
      "step": 40
    },
    {
      "epoch": 0.02,
      "learning_rate": 9.520613614573348e-05,
      "loss": 0.4806,
      "step": 50
    },
    {
      "epoch": 0.03,
      "learning_rate": 9.424736337488016e-05,
      "loss": 0.4198,
      "step": 60
    },
    {
      "epoch": 0.03,
      "learning_rate": 9.328859060402684e-05,
      "loss": 0.414,
      "step": 70
    },
    {
      "epoch": 0.04,
      "learning_rate": 9.232981783317354e-05,
      "loss": 0.3727,
      "step": 80
    },
    {
      "epoch": 0.04,
      "learning_rate": 9.137104506232024e-05,
      "loss": 0.3566,
      "step": 90
    },
    {
      "epoch": 0.05,
      "learning_rate": 9.041227229146692e-05,
      "loss": 0.3575,
      "step": 100
    },
    {
      "epoch": 0.05,
      "learning_rate": 8.945349952061362e-05,
      "loss": 0.3312,
      "step": 110
    },
    {
      "epoch": 0.06,
      "learning_rate": 8.849472674976032e-05,
      "loss": 0.3349,
      "step": 120
    },
    {
      "epoch": 0.06,
      "learning_rate": 8.7535953978907e-05,
      "loss": 0.3095,
      "step": 130
    },
    {
      "epoch": 0.07,
      "learning_rate": 8.65771812080537e-05,
      "loss": 0.3227,
      "step": 140
    },
    {
      "epoch": 0.07,
      "learning_rate": 8.56184084372004e-05,
      "loss": 0.3074,
      "step": 150
    },
    {
      "epoch": 0.08,
      "learning_rate": 8.465963566634708e-05,
      "loss": 0.2834,
      "step": 160
    },
    {
      "epoch": 0.08,
      "learning_rate": 8.370086289549378e-05,
      "loss": 0.2855,
      "step": 170
    },
    {
      "epoch": 0.09,
      "learning_rate": 8.274209012464046e-05,
      "loss": 0.2981,
      "step": 180
    },
    {
      "epoch": 0.09,
      "learning_rate": 8.178331735378716e-05,
      "loss": 0.2834,
      "step": 190
    },
    {
      "epoch": 0.1,
      "learning_rate": 8.082454458293385e-05,
      "loss": 0.264,
      "step": 200
    },
    {
      "epoch": 0.1,
      "learning_rate": 7.986577181208054e-05,
      "loss": 0.2732,
      "step": 210
    },
    {
      "epoch": 0.11,
      "learning_rate": 7.890699904122723e-05,
      "loss": 0.2758,
      "step": 220
    },
    {
      "epoch": 0.11,
      "learning_rate": 7.794822627037393e-05,
      "loss": 0.2626,
      "step": 230
    },
    {
      "epoch": 0.12,
      "learning_rate": 7.698945349952061e-05,
      "loss": 0.2516,
      "step": 240
    },
    {
      "epoch": 0.12,
      "learning_rate": 7.603068072866731e-05,
      "loss": 0.2575,
      "step": 250
    },
    {
      "epoch": 0.12,
      "learning_rate": 7.507190795781401e-05,
      "loss": 0.2565,
      "step": 260
    },
    {
      "epoch": 0.13,
      "learning_rate": 7.411313518696069e-05,
      "loss": 0.2382,
      "step": 270
    },
    {
      "epoch": 0.13,
      "learning_rate": 7.315436241610739e-05,
      "loss": 0.2489,
      "step": 280
    },
    {
      "epoch": 0.14,
      "learning_rate": 7.219558964525409e-05,
      "loss": 0.2308,
      "step": 290
    },
    {
      "epoch": 0.14,
      "learning_rate": 7.123681687440077e-05,
      "loss": 0.2299,
      "step": 300
    },
    {
      "epoch": 0.15,
      "learning_rate": 7.027804410354745e-05,
      "loss": 0.2479,
      "step": 310
    },
    {
      "epoch": 0.15,
      "learning_rate": 6.931927133269415e-05,
      "loss": 0.2288,
      "step": 320
    },
    {
      "epoch": 0.16,
      "learning_rate": 6.836049856184085e-05,
      "loss": 0.2394,
      "step": 330
    },
    {
      "epoch": 0.16,
      "learning_rate": 6.740172579098753e-05,
      "loss": 0.2239,
      "step": 340
    },
    {
      "epoch": 0.17,
      "learning_rate": 6.644295302013423e-05,
      "loss": 0.2282,
      "step": 350
    },
    {
      "epoch": 0.17,
      "learning_rate": 6.548418024928093e-05,
      "loss": 0.2227,
      "step": 360
    },
    {
      "epoch": 0.18,
      "learning_rate": 6.452540747842761e-05,
      "loss": 0.2181,
      "step": 370
    },
    {
      "epoch": 0.18,
      "learning_rate": 6.356663470757431e-05,
      "loss": 0.2211,
      "step": 380
    },
    {
      "epoch": 0.19,
      "learning_rate": 6.2607861936721e-05,
      "loss": 0.2157,
      "step": 390
    },
    {
      "epoch": 0.19,
      "learning_rate": 6.164908916586769e-05,
      "loss": 0.2097,
      "step": 400
    },
    {
      "epoch": 0.2,
      "learning_rate": 6.069031639501438e-05,
      "loss": 0.2006,
      "step": 410
    },
    {
      "epoch": 0.2,
      "learning_rate": 5.973154362416108e-05,
      "loss": 0.1968,
      "step": 420
    },
    {
      "epoch": 0.21,
      "learning_rate": 5.8772770853307766e-05,
      "loss": 0.2052,
      "step": 430
    },
    {
      "epoch": 0.21,
      "learning_rate": 5.7813998082454456e-05,
      "loss": 0.1958,
      "step": 440
    },
    {
      "epoch": 0.22,
      "learning_rate": 5.6855225311601154e-05,
      "loss": 0.1923,
      "step": 450
    },
    {
      "epoch": 0.22,
      "learning_rate": 5.5896452540747844e-05,
      "loss": 0.1968,
      "step": 460
    },
    {
      "epoch": 0.23,
      "learning_rate": 5.4937679769894534e-05,
      "loss": 0.1862,
      "step": 470
    },
    {
      "epoch": 0.23,
      "learning_rate": 5.397890699904123e-05,
      "loss": 0.1957,
      "step": 480
    },
    {
      "epoch": 0.23,
      "learning_rate": 5.302013422818792e-05,
      "loss": 0.1863,
      "step": 490
    },
    {
      "epoch": 0.24,
      "learning_rate": 5.206136145733461e-05,
      "loss": 0.1894,
      "step": 500
    },
    {
      "epoch": 0.24,
      "learning_rate": 5.110258868648131e-05,
      "loss": 0.1837,
      "step": 510
    },
    {
      "epoch": 0.25,
      "learning_rate": 5.0143815915628e-05,
      "loss": 0.1836,
      "step": 520
    },
    {
      "epoch": 0.25,
      "learning_rate": 4.918504314477469e-05,
      "loss": 0.1871,
      "step": 530
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.822627037392138e-05,
      "loss": 0.1795,
      "step": 540
    },
    {
      "epoch": 0.26,
      "learning_rate": 4.726749760306807e-05,
      "loss": 0.1846,
      "step": 550
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.630872483221477e-05,
      "loss": 0.1825,
      "step": 560
    },
    {
      "epoch": 0.27,
      "learning_rate": 4.534995206136146e-05,
      "loss": 0.1704,
      "step": 570
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.439117929050815e-05,
      "loss": 0.1781,
      "step": 580
    },
    {
      "epoch": 0.28,
      "learning_rate": 4.3432406519654846e-05,
      "loss": 0.1773,
      "step": 590
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.2473633748801537e-05,
      "loss": 0.1742,
      "step": 600
    },
    {
      "epoch": 0.29,
      "learning_rate": 4.151486097794823e-05,
      "loss": 0.1683,
      "step": 610
    },
    {
      "epoch": 0.3,
      "learning_rate": 4.0556088207094924e-05,
      "loss": 0.1634,
      "step": 620
    },
    {
      "epoch": 0.3,
      "learning_rate": 3.959731543624161e-05,
      "loss": 0.161,
      "step": 630
    },
    {
      "epoch": 0.31,
      "learning_rate": 3.8638542665388305e-05,
      "loss": 0.166,
      "step": 640
    },
    {
      "epoch": 0.31,
      "learning_rate": 3.7679769894535e-05,
      "loss": 0.1689,
      "step": 650
    },
    {
      "epoch": 0.32,
      "learning_rate": 3.6720997123681686e-05,
      "loss": 0.1726,
      "step": 660
    },
    {
      "epoch": 0.32,
      "learning_rate": 3.576222435282838e-05,
      "loss": 0.1509,
      "step": 670
    },
    {
      "epoch": 0.33,
      "learning_rate": 3.480345158197507e-05,
      "loss": 0.1588,
      "step": 680
    },
    {
      "epoch": 0.33,
      "learning_rate": 3.3844678811121764e-05,
      "loss": 0.1558,
      "step": 690
    },
    {
      "epoch": 0.34,
      "learning_rate": 3.288590604026846e-05,
      "loss": 0.1522,
      "step": 700
    },
    {
      "epoch": 0.34,
      "learning_rate": 3.192713326941515e-05,
      "loss": 0.1568,
      "step": 710
    },
    {
      "epoch": 0.35,
      "learning_rate": 3.096836049856184e-05,
      "loss": 0.1471,
      "step": 720
    },
    {
      "epoch": 0.35,
      "learning_rate": 3.0009587727708532e-05,
      "loss": 0.1437,
      "step": 730
    },
    {
      "epoch": 0.35,
      "learning_rate": 2.9050814956855226e-05,
      "loss": 0.1507,
      "step": 740
    },
    {
      "epoch": 0.36,
      "learning_rate": 2.8092042186001923e-05,
      "loss": 0.1398,
      "step": 750
    },
    {
      "epoch": 0.36,
      "learning_rate": 2.713326941514861e-05,
      "loss": 0.1429,
      "step": 760
    },
    {
      "epoch": 0.37,
      "learning_rate": 2.6174496644295304e-05,
      "loss": 0.151,
      "step": 770
    },
    {
      "epoch": 0.37,
      "learning_rate": 2.5215723873441994e-05,
      "loss": 0.1419,
      "step": 780
    },
    {
      "epoch": 0.38,
      "learning_rate": 2.4256951102588688e-05,
      "loss": 0.1482,
      "step": 790
    },
    {
      "epoch": 0.38,
      "learning_rate": 2.3298178331735378e-05,
      "loss": 0.1395,
      "step": 800
    },
    {
      "epoch": 0.39,
      "learning_rate": 2.2339405560882072e-05,
      "loss": 0.1465,
      "step": 810
    },
    {
      "epoch": 0.39,
      "learning_rate": 2.1380632790028766e-05,
      "loss": 0.144,
      "step": 820
    },
    {
      "epoch": 0.4,
      "learning_rate": 2.0421860019175456e-05,
      "loss": 0.1484,
      "step": 830
    },
    {
      "epoch": 0.4,
      "learning_rate": 1.946308724832215e-05,
      "loss": 0.1367,
      "step": 840
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.850431447746884e-05,
      "loss": 0.1342,
      "step": 850
    },
    {
      "epoch": 0.41,
      "learning_rate": 1.7545541706615534e-05,
      "loss": 0.1293,
      "step": 860
    },
    {
      "epoch": 0.42,
      "learning_rate": 1.6586768935762224e-05,
      "loss": 0.1346,
      "step": 870
    },
    {
      "epoch": 0.42,
      "learning_rate": 1.5627996164908918e-05,
      "loss": 0.1351,
      "step": 880
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.4669223394055609e-05,
      "loss": 0.1289,
      "step": 890
    },
    {
      "epoch": 0.43,
      "learning_rate": 1.37104506232023e-05,
      "loss": 0.1312,
      "step": 900
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.2751677852348994e-05,
      "loss": 0.1309,
      "step": 910
    },
    {
      "epoch": 0.44,
      "learning_rate": 1.1792905081495687e-05,
      "loss": 0.1225,
      "step": 920
    },
    {
      "epoch": 0.45,
      "learning_rate": 1.0834132310642379e-05,
      "loss": 0.1255,
      "step": 930
    },
    {
      "epoch": 0.45,
      "learning_rate": 9.87535953978907e-06,
      "loss": 0.1341,
      "step": 940
    },
    {
      "epoch": 0.46,
      "learning_rate": 8.916586768935763e-06,
      "loss": 0.1299,
      "step": 950
    },
    {
      "epoch": 0.46,
      "learning_rate": 7.957813998082455e-06,
      "loss": 0.1346,
      "step": 960
    },
    {
      "epoch": 0.46,
      "learning_rate": 6.999041227229147e-06,
      "loss": 0.1261,
      "step": 970
    },
    {
      "epoch": 0.47,
      "learning_rate": 6.04026845637584e-06,
      "loss": 0.1239,
      "step": 980
    },
    {
      "epoch": 0.47,
      "learning_rate": 5.081495685522531e-06,
      "loss": 0.1185,
      "step": 990
    },
    {
      "epoch": 0.48,
      "learning_rate": 4.122722914669223e-06,
      "loss": 0.1228,
      "step": 1000
    },
    {
      "epoch": 0.48,
      "learning_rate": 3.1639501438159157e-06,
      "loss": 0.1243,
      "step": 1010
    },
    {
      "epoch": 0.49,
      "learning_rate": 2.2051773729626078e-06,
      "loss": 0.1243,
      "step": 1020
    },
    {
      "epoch": 0.49,
      "learning_rate": 1.2464046021093003e-06,
      "loss": 0.1259,
      "step": 1030
    },
    {
      "epoch": 0.5,
      "learning_rate": 2.876318312559924e-07,
      "loss": 0.1258,
      "step": 1040
    },
    {
      "epoch": 0.5,
      "step": 1043,
      "total_flos": 3.162519559471104e+16,
      "train_loss": 0.23776370769836302,
      "train_runtime": 7748.8771,
      "train_samples_per_second": 68.929,
      "train_steps_per_second": 0.135
    }
  ],
  "logging_steps": 10,
  "max_steps": 1043,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 3.162519559471104e+16,
  "trial_name": null,
  "trial_params": null
}