{
  "best_metric": 0.616822429906542,
  "best_model_checkpoint": "SW2-RHS-DA\\checkpoint-292",
  "epoch": 39.111111111111114,
  "eval_steps": 500,
  "global_step": 880,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.44,
      "learning_rate": 3.954545454545455e-05,
      "loss": 1.6539,
      "step": 10
    },
    {
      "epoch": 0.89,
      "learning_rate": 3.909090909090909e-05,
      "loss": 1.302,
      "step": 20
    },
    {
      "epoch": 0.98,
      "eval_accuracy": 0.411214953271028,
      "eval_loss": 2.847358226776123,
      "eval_runtime": 2.4491,
      "eval_samples_per_second": 43.69,
      "eval_steps_per_second": 2.858,
      "step": 22
    },
    {
      "epoch": 1.33,
      "learning_rate": 3.863636363636364e-05,
      "loss": 0.8548,
      "step": 30
    },
    {
      "epoch": 1.78,
      "learning_rate": 3.818181818181819e-05,
      "loss": 0.6224,
      "step": 40
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.411214953271028,
      "eval_loss": 1.4649972915649414,
      "eval_runtime": 2.5381,
      "eval_samples_per_second": 42.157,
      "eval_steps_per_second": 2.758,
      "step": 45
    },
    {
      "epoch": 2.22,
      "learning_rate": 3.772727272727273e-05,
      "loss": 0.4983,
      "step": 50
    },
    {
      "epoch": 2.67,
      "learning_rate": 3.7272727272727276e-05,
      "loss": 0.3905,
      "step": 60
    },
    {
      "epoch": 2.98,
      "eval_accuracy": 0.411214953271028,
      "eval_loss": 1.486470103263855,
      "eval_runtime": 2.5569,
      "eval_samples_per_second": 41.847,
      "eval_steps_per_second": 2.738,
      "step": 67
    },
    {
      "epoch": 3.11,
      "learning_rate": 3.681818181818182e-05,
      "loss": 0.3143,
      "step": 70
    },
    {
      "epoch": 3.56,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 0.2038,
      "step": 80
    },
    {
      "epoch": 4.0,
      "learning_rate": 3.590909090909091e-05,
      "loss": 0.1416,
      "step": 90
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5981308411214953,
      "eval_loss": 0.9452531933784485,
      "eval_runtime": 2.8177,
      "eval_samples_per_second": 37.974,
      "eval_steps_per_second": 2.484,
      "step": 90
    },
    {
      "epoch": 4.44,
      "learning_rate": 3.545454545454546e-05,
      "loss": 0.1267,
      "step": 100
    },
    {
      "epoch": 4.89,
      "learning_rate": 3.5000000000000004e-05,
      "loss": 0.1116,
      "step": 110
    },
    {
      "epoch": 4.98,
      "eval_accuracy": 0.5514018691588785,
      "eval_loss": 0.9800727963447571,
      "eval_runtime": 2.4796,
      "eval_samples_per_second": 43.152,
      "eval_steps_per_second": 2.823,
      "step": 112
    },
    {
      "epoch": 5.33,
      "learning_rate": 3.454545454545455e-05,
      "loss": 0.1516,
      "step": 120
    },
    {
      "epoch": 5.78,
      "learning_rate": 3.409090909090909e-05,
      "loss": 0.0866,
      "step": 130
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.6074766355140186,
      "eval_loss": 1.520142912864685,
      "eval_runtime": 2.5261,
      "eval_samples_per_second": 42.358,
      "eval_steps_per_second": 2.771,
      "step": 135
    },
    {
      "epoch": 6.22,
      "learning_rate": 3.363636363636364e-05,
      "loss": 0.1082,
      "step": 140
    },
    {
      "epoch": 6.67,
      "learning_rate": 3.318181818181819e-05,
      "loss": 0.0579,
      "step": 150
    },
    {
      "epoch": 6.98,
      "eval_accuracy": 0.5981308411214953,
      "eval_loss": 1.723357915878296,
      "eval_runtime": 2.4956,
      "eval_samples_per_second": 42.876,
      "eval_steps_per_second": 2.805,
      "step": 157
    },
    {
      "epoch": 7.11,
      "learning_rate": 3.272727272727273e-05,
      "loss": 0.0679,
      "step": 160
    },
    {
      "epoch": 7.56,
      "learning_rate": 3.2272727272727276e-05,
      "loss": 0.1047,
      "step": 170
    },
    {
      "epoch": 8.0,
      "learning_rate": 3.181818181818182e-05,
      "loss": 0.0667,
      "step": 180
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.5981308411214953,
      "eval_loss": 1.9649068117141724,
      "eval_runtime": 2.4861,
      "eval_samples_per_second": 43.04,
      "eval_steps_per_second": 2.816,
      "step": 180
    },
    {
      "epoch": 8.44,
      "learning_rate": 3.1363636363636365e-05,
      "loss": 0.0661,
      "step": 190
    },
    {
      "epoch": 8.89,
      "learning_rate": 3.090909090909091e-05,
      "loss": 0.0664,
      "step": 200
    },
    {
      "epoch": 8.98,
      "eval_accuracy": 0.5981308411214953,
      "eval_loss": 1.9504597187042236,
      "eval_runtime": 2.4306,
      "eval_samples_per_second": 44.022,
      "eval_steps_per_second": 2.88,
      "step": 202
    },
    {
      "epoch": 9.33,
      "learning_rate": 3.0454545454545456e-05,
      "loss": 0.0647,
      "step": 210
    },
    {
      "epoch": 9.78,
      "learning_rate": 3.0000000000000004e-05,
      "loss": 0.0742,
      "step": 220
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.5981308411214953,
      "eval_loss": 1.9448089599609375,
      "eval_runtime": 2.4836,
      "eval_samples_per_second": 43.083,
      "eval_steps_per_second": 2.819,
      "step": 225
    },
    {
      "epoch": 10.22,
      "learning_rate": 2.954545454545455e-05,
      "loss": 0.0588,
      "step": 230
    },
    {
      "epoch": 10.67,
      "learning_rate": 2.9090909090909093e-05,
      "loss": 0.0558,
      "step": 240
    },
    {
      "epoch": 10.98,
      "eval_accuracy": 0.5981308411214953,
      "eval_loss": 1.9545286893844604,
      "eval_runtime": 2.5206,
      "eval_samples_per_second": 42.45,
      "eval_steps_per_second": 2.777,
      "step": 247
    },
    {
      "epoch": 11.11,
      "learning_rate": 2.863636363636364e-05,
      "loss": 0.0699,
      "step": 250
    },
    {
      "epoch": 11.56,
      "learning_rate": 2.8181818181818185e-05,
      "loss": 0.0564,
      "step": 260
    },
    {
      "epoch": 12.0,
      "learning_rate": 2.7727272727272732e-05,
      "loss": 0.0475,
      "step": 270
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.5887850467289719,
      "eval_loss": 2.1516401767730713,
      "eval_runtime": 2.4679,
      "eval_samples_per_second": 43.356,
      "eval_steps_per_second": 2.836,
      "step": 270
    },
    {
      "epoch": 12.44,
      "learning_rate": 2.7272727272727273e-05,
      "loss": 0.0408,
      "step": 280
    },
    {
      "epoch": 12.89,
      "learning_rate": 2.6818181818181817e-05,
      "loss": 0.114,
      "step": 290
    },
    {
      "epoch": 12.98,
      "eval_accuracy": 0.616822429906542,
      "eval_loss": 2.1001620292663574,
      "eval_runtime": 2.4814,
      "eval_samples_per_second": 43.122,
      "eval_steps_per_second": 2.821,
      "step": 292
    },
    {
      "epoch": 13.33,
      "learning_rate": 2.6363636363636365e-05,
      "loss": 0.0592,
      "step": 300
    },
    {
      "epoch": 13.78,
      "learning_rate": 2.590909090909091e-05,
      "loss": 0.051,
      "step": 310
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.5981308411214953,
      "eval_loss": 2.2643392086029053,
      "eval_runtime": 2.4411,
      "eval_samples_per_second": 43.833,
      "eval_steps_per_second": 2.868,
      "step": 315
    },
    {
      "epoch": 14.22,
      "learning_rate": 2.5454545454545457e-05,
      "loss": 0.0575,
      "step": 320
    },
    {
      "epoch": 14.67,
      "learning_rate": 2.5e-05,
      "loss": 0.0318,
      "step": 330
    },
    {
      "epoch": 14.98,
      "eval_accuracy": 0.5981308411214953,
      "eval_loss": 2.346831798553467,
      "eval_runtime": 2.4771,
      "eval_samples_per_second": 43.196,
      "eval_steps_per_second": 2.826,
      "step": 337
    },
    {
      "epoch": 15.11,
      "learning_rate": 2.454545454545455e-05,
      "loss": 0.0273,
      "step": 340
    },
    {
      "epoch": 15.56,
      "learning_rate": 2.4090909090909093e-05,
      "loss": 0.0232,
      "step": 350
    },
    {
      "epoch": 16.0,
      "learning_rate": 2.363636363636364e-05,
      "loss": 0.0673,
      "step": 360
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.6074766355140186,
      "eval_loss": 2.334057092666626,
      "eval_runtime": 2.5831,
      "eval_samples_per_second": 41.422,
      "eval_steps_per_second": 2.71,
      "step": 360
    },
    {
      "epoch": 16.44,
      "learning_rate": 2.3181818181818185e-05,
      "loss": 0.0402,
      "step": 370
    },
    {
      "epoch": 16.89,
      "learning_rate": 2.2727272727272733e-05,
      "loss": 0.0566,
      "step": 380
    },
    {
      "epoch": 16.98,
      "eval_accuracy": 0.6074766355140186,
      "eval_loss": 2.319139003753662,
      "eval_runtime": 2.4296,
      "eval_samples_per_second": 44.04,
      "eval_steps_per_second": 2.881,
      "step": 382
    },
    {
      "epoch": 17.33,
      "learning_rate": 2.2272727272727274e-05,
      "loss": 0.0445,
      "step": 390
    },
    {
      "epoch": 17.78,
      "learning_rate": 2.1818181818181818e-05,
      "loss": 0.0543,
      "step": 400
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.616822429906542,
      "eval_loss": 2.280747413635254,
      "eval_runtime": 2.5226,
      "eval_samples_per_second": 42.417,
      "eval_steps_per_second": 2.775,
      "step": 405
    },
    {
      "epoch": 18.22,
      "learning_rate": 2.1363636363636365e-05,
      "loss": 0.0419,
      "step": 410
    },
    {
      "epoch": 18.67,
      "learning_rate": 2.090909090909091e-05,
      "loss": 0.0458,
      "step": 420
    },
    {
      "epoch": 18.98,
      "eval_accuracy": 0.616822429906542,
      "eval_loss": 2.2148241996765137,
      "eval_runtime": 2.5536,
      "eval_samples_per_second": 41.901,
      "eval_steps_per_second": 2.741,
      "step": 427
    },
    {
      "epoch": 19.11,
      "learning_rate": 2.0454545454545457e-05,
      "loss": 0.0255,
      "step": 430
    },
    {
      "epoch": 19.56,
      "learning_rate": 2e-05,
      "loss": 0.0306,
      "step": 440
    },
    {
      "epoch": 20.0,
      "learning_rate": 1.9545454545454546e-05,
      "loss": 0.0421,
      "step": 450
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.5887850467289719,
      "eval_loss": 2.5467782020568848,
      "eval_runtime": 2.4846,
      "eval_samples_per_second": 43.065,
      "eval_steps_per_second": 2.817,
      "step": 450
    },
    {
      "epoch": 20.44,
      "learning_rate": 1.9090909090909094e-05,
      "loss": 0.0509,
      "step": 460
    },
    {
      "epoch": 20.89,
      "learning_rate": 1.8636363636363638e-05,
      "loss": 0.0139,
      "step": 470
    },
    {
      "epoch": 20.98,
      "eval_accuracy": 0.616822429906542,
      "eval_loss": 2.2408316135406494,
      "eval_runtime": 2.5411,
      "eval_samples_per_second": 42.108,
      "eval_steps_per_second": 2.755,
      "step": 472
    },
    {
      "epoch": 21.33,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.0208,
      "step": 480
    },
    {
      "epoch": 21.78,
      "learning_rate": 1.772727272727273e-05,
      "loss": 0.012,
      "step": 490
    },
    {
      "epoch": 22.0,
      "eval_accuracy": 0.5887850467289719,
      "eval_loss": 2.568880796432495,
      "eval_runtime": 2.4491,
      "eval_samples_per_second": 43.69,
      "eval_steps_per_second": 2.858,
      "step": 495
    },
    {
      "epoch": 22.22,
      "learning_rate": 1.7272727272727274e-05,
      "loss": 0.0375,
      "step": 500
    },
    {
      "epoch": 22.67,
      "learning_rate": 1.681818181818182e-05,
      "loss": 0.017,
      "step": 510
    },
    {
      "epoch": 22.98,
      "eval_accuracy": 0.616822429906542,
      "eval_loss": 2.548487901687622,
      "eval_runtime": 2.5416,
      "eval_samples_per_second": 42.1,
      "eval_steps_per_second": 2.754,
      "step": 517
    },
    {
      "epoch": 23.11,
      "learning_rate": 1.6363636363636366e-05,
      "loss": 0.0293,
      "step": 520
    },
    {
      "epoch": 23.56,
      "learning_rate": 1.590909090909091e-05,
      "loss": 0.0482,
      "step": 530
    },
    {
      "epoch": 24.0,
      "learning_rate": 1.5454545454545454e-05,
      "loss": 0.0529,
      "step": 540
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.6074766355140186,
      "eval_loss": 2.674633502960205,
      "eval_runtime": 2.5767,
      "eval_samples_per_second": 41.526,
      "eval_steps_per_second": 2.717,
      "step": 540
    },
    {
      "epoch": 24.44,
      "learning_rate": 1.5000000000000002e-05,
      "loss": 0.0063,
      "step": 550
    },
    {
      "epoch": 24.89,
      "learning_rate": 1.4545454545454546e-05,
      "loss": 0.0414,
      "step": 560
    },
    {
      "epoch": 24.98,
      "eval_accuracy": 0.5887850467289719,
      "eval_loss": 2.7693288326263428,
      "eval_runtime": 2.6239,
      "eval_samples_per_second": 40.778,
      "eval_steps_per_second": 2.668,
      "step": 562
    },
    {
      "epoch": 25.33,
      "learning_rate": 1.4090909090909092e-05,
      "loss": 0.0203,
      "step": 570
    },
    {
      "epoch": 25.78,
      "learning_rate": 1.3636363636363637e-05,
      "loss": 0.0158,
      "step": 580
    },
    {
      "epoch": 26.0,
      "eval_accuracy": 0.5887850467289719,
      "eval_loss": 2.7446506023406982,
      "eval_runtime": 2.5651,
      "eval_samples_per_second": 41.714,
      "eval_steps_per_second": 2.729,
      "step": 585
    },
    {
      "epoch": 26.22,
      "learning_rate": 1.3181818181818183e-05,
      "loss": 0.0185,
      "step": 590
    },
    {
      "epoch": 26.67,
      "learning_rate": 1.2727272727272728e-05,
      "loss": 0.0205,
      "step": 600
    },
    {
      "epoch": 26.98,
      "eval_accuracy": 0.5887850467289719,
      "eval_loss": 2.8566343784332275,
      "eval_runtime": 2.5211,
      "eval_samples_per_second": 42.442,
      "eval_steps_per_second": 2.777,
      "step": 607
    },
    {
      "epoch": 27.11,
      "learning_rate": 1.2272727272727274e-05,
      "loss": 0.04,
      "step": 610
    },
    {
      "epoch": 27.56,
      "learning_rate": 1.181818181818182e-05,
      "loss": 0.0133,
      "step": 620
    },
    {
      "epoch": 28.0,
      "learning_rate": 1.1363636363636366e-05,
      "loss": 0.0205,
      "step": 630
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.5887850467289719,
      "eval_loss": 2.846874713897705,
      "eval_runtime": 2.4371,
      "eval_samples_per_second": 43.905,
      "eval_steps_per_second": 2.872,
      "step": 630
    },
    {
      "epoch": 28.44,
      "learning_rate": 1.0909090909090909e-05,
      "loss": 0.0211,
      "step": 640
    },
    {
      "epoch": 28.89,
      "learning_rate": 1.0454545454545455e-05,
      "loss": 0.006,
      "step": 650
    },
    {
      "epoch": 28.98,
      "eval_accuracy": 0.5887850467289719,
      "eval_loss": 2.9508450031280518,
      "eval_runtime": 2.4616,
      "eval_samples_per_second": 43.468,
      "eval_steps_per_second": 2.844,
      "step": 652
    },
    {
      "epoch": 29.33,
      "learning_rate": 1e-05,
      "loss": 0.0198,
      "step": 660
    },
    {
      "epoch": 29.78,
      "learning_rate": 9.545454545454547e-06,
      "loss": 0.0061,
      "step": 670
    },
    {
      "epoch": 30.0,
      "eval_accuracy": 0.5887850467289719,
      "eval_loss": 3.0560247898101807,
      "eval_runtime": 2.6005,
      "eval_samples_per_second": 41.145,
      "eval_steps_per_second": 2.692,
      "step": 675
    },
    {
      "epoch": 30.22,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.0124,
      "step": 680
    },
    {
      "epoch": 30.67,
      "learning_rate": 8.636363636363637e-06,
      "loss": 0.0227,
      "step": 690
    },
    {
      "epoch": 30.98,
      "eval_accuracy": 0.5887850467289719,
      "eval_loss": 3.0431346893310547,
      "eval_runtime": 2.5396,
      "eval_samples_per_second": 42.132,
      "eval_steps_per_second": 2.756,
      "step": 697
    },
    {
      "epoch": 31.11,
      "learning_rate": 8.181818181818183e-06,
      "loss": 0.0261,
      "step": 700
    },
    {
      "epoch": 31.56,
      "learning_rate": 7.727272727272727e-06,
      "loss": 0.0174,
      "step": 710
    },
    {
      "epoch": 32.0,
      "learning_rate": 7.272727272727273e-06,
      "loss": 0.034,
      "step": 720
    },
    {
      "epoch": 32.0,
      "eval_accuracy": 0.5887850467289719,
      "eval_loss": 3.0496559143066406,
      "eval_runtime": 2.4746,
      "eval_samples_per_second": 43.24,
      "eval_steps_per_second": 2.829,
      "step": 720
    },
    {
      "epoch": 32.44,
      "learning_rate": 6.818181818181818e-06,
      "loss": 0.0277,
      "step": 730
    },
    {
      "epoch": 32.89,
      "learning_rate": 6.363636363636364e-06,
      "loss": 0.0039,
      "step": 740
    },
    {
      "epoch": 32.98,
      "eval_accuracy": 0.5887850467289719,
      "eval_loss": 3.0935721397399902,
      "eval_runtime": 2.5656,
      "eval_samples_per_second": 41.706,
      "eval_steps_per_second": 2.728,
      "step": 742
    },
    {
      "epoch": 33.33,
      "learning_rate": 5.90909090909091e-06,
      "loss": 0.02,
      "step": 750
    },
    {
      "epoch": 33.78,
      "learning_rate": 5.4545454545454545e-06,
      "loss": 0.0031,
      "step": 760
    },
    {
      "epoch": 34.0,
      "eval_accuracy": 0.5887850467289719,
      "eval_loss": 3.1158316135406494,
      "eval_runtime": 2.4251,
      "eval_samples_per_second": 44.123,
      "eval_steps_per_second": 2.887,
      "step": 765
    },
    {
      "epoch": 34.22,
      "learning_rate": 5e-06,
      "loss": 0.0178,
      "step": 770
    },
    {
      "epoch": 34.67,
      "learning_rate": 4.5454545454545455e-06,
      "loss": 0.0118,
      "step": 780
    },
    {
      "epoch": 34.98,
      "eval_accuracy": 0.5887850467289719,
      "eval_loss": 3.109339714050293,
      "eval_runtime": 2.5796,
      "eval_samples_per_second": 41.479,
      "eval_steps_per_second": 2.714,
      "step": 787
    },
    {
      "epoch": 35.11,
      "learning_rate": 4.0909090909090915e-06,
      "loss": 0.0061,
      "step": 790
    },
    {
      "epoch": 35.56,
      "learning_rate": 3.6363636363636366e-06,
      "loss": 0.018,
      "step": 800
    },
    {
      "epoch": 36.0,
      "learning_rate": 3.181818181818182e-06,
      "loss": 0.0045,
      "step": 810
    },
    {
      "epoch": 36.0,
      "eval_accuracy": 0.5887850467289719,
      "eval_loss": 3.088502883911133,
      "eval_runtime": 2.4691,
      "eval_samples_per_second": 43.336,
      "eval_steps_per_second": 2.835,
      "step": 810
    },
    {
      "epoch": 36.44,
      "learning_rate": 2.7272727272727272e-06,
      "loss": 0.0224,
      "step": 820
    },
    {
      "epoch": 36.89,
      "learning_rate": 2.2727272727272728e-06,
      "loss": 0.0168,
      "step": 830
    },
    {
      "epoch": 36.98,
      "eval_accuracy": 0.5887850467289719,
      "eval_loss": 3.0813817977905273,
      "eval_runtime": 2.46,
      "eval_samples_per_second": 43.495,
      "eval_steps_per_second": 2.845,
      "step": 832
    },
    {
      "epoch": 37.33,
      "learning_rate": 1.8181818181818183e-06,
      "loss": 0.0106,
      "step": 840
    },
    {
      "epoch": 37.78,
      "learning_rate": 1.3636363636363636e-06,
      "loss": 0.0071,
      "step": 850
    },
    {
      "epoch": 38.0,
      "eval_accuracy": 0.5887850467289719,
      "eval_loss": 3.073270320892334,
      "eval_runtime": 2.8765,
      "eval_samples_per_second": 37.198,
      "eval_steps_per_second": 2.433,
      "step": 855
    },
    {
      "epoch": 38.22,
      "learning_rate": 9.090909090909091e-07,
      "loss": 0.0123,
      "step": 860
    },
    {
      "epoch": 38.67,
      "learning_rate": 4.5454545454545457e-07,
      "loss": 0.0238,
      "step": 870
    },
    {
      "epoch": 38.98,
      "eval_accuracy": 0.5887850467289719,
      "eval_loss": 3.0825016498565674,
      "eval_runtime": 2.6096,
      "eval_samples_per_second": 41.002,
      "eval_steps_per_second": 2.682,
      "step": 877
    },
    {
      "epoch": 39.11,
      "learning_rate": 0.0,
      "loss": 0.0072,
      "step": 880
    },
    {
      "epoch": 39.11,
      "eval_accuracy": 0.5887850467289719,
      "eval_loss": 3.082833766937256,
      "eval_runtime": 2.596,
      "eval_samples_per_second": 41.217,
      "eval_steps_per_second": 2.696,
      "step": 880
    },
    {
      "epoch": 39.11,
      "step": 880,
      "total_flos": 1.8272580263466762e+18,
      "train_loss": 0.10380522197493437,
      "train_runtime": 1214.1801,
      "train_samples_per_second": 47.308,
      "train_steps_per_second": 0.725
    }
  ],
  "logging_steps": 10,
  "max_steps": 880,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 40,
  "save_steps": 500,
  "total_flos": 1.8272580263466762e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}