{
  "best_metric": 0.9193548387096774,
  "best_model_checkpoint": "SW2-TO-DA\\checkpoint-130",
  "epoch": 38.62068965517241,
  "eval_steps": 500,
  "global_step": 560,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.69,
      "learning_rate": 2.6785714285714284e-05,
      "loss": 1.4955,
      "step": 10
    },
    {
      "epoch": 0.97,
      "eval_accuracy": 0.08064516129032258,
      "eval_loss": 1.558031439781189,
      "eval_runtime": 2.4051,
      "eval_samples_per_second": 25.778,
      "eval_steps_per_second": 1.663,
      "step": 14
    },
    {
      "epoch": 1.38,
      "learning_rate": 5.357142857142857e-05,
      "loss": 1.3943,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.6451612903225806,
      "eval_loss": 1.1315828561782837,
      "eval_runtime": 1.9485,
      "eval_samples_per_second": 31.82,
      "eval_steps_per_second": 2.053,
      "step": 29
    },
    {
      "epoch": 2.07,
      "learning_rate": 8.035714285714285e-05,
      "loss": 1.2678,
      "step": 30
    },
    {
      "epoch": 2.76,
      "learning_rate": 0.00010714285714285714,
      "loss": 1.0056,
      "step": 40
    },
    {
      "epoch": 2.97,
      "eval_accuracy": 0.7419354838709677,
      "eval_loss": 0.6406522393226624,
      "eval_runtime": 2.1315,
      "eval_samples_per_second": 29.087,
      "eval_steps_per_second": 1.877,
      "step": 43
    },
    {
      "epoch": 3.45,
      "learning_rate": 0.00013392857142857144,
      "loss": 0.7744,
      "step": 50
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8709677419354839,
      "eval_loss": 0.4265057146549225,
      "eval_runtime": 1.9565,
      "eval_samples_per_second": 31.69,
      "eval_steps_per_second": 2.045,
      "step": 58
    },
    {
      "epoch": 4.14,
      "learning_rate": 0.0001488095238095238,
      "loss": 0.7109,
      "step": 60
    },
    {
      "epoch": 4.83,
      "learning_rate": 0.00014583333333333332,
      "loss": 0.6022,
      "step": 70
    },
    {
      "epoch": 4.97,
      "eval_accuracy": 0.8548387096774194,
      "eval_loss": 0.4360726773738861,
      "eval_runtime": 1.949,
      "eval_samples_per_second": 31.812,
      "eval_steps_per_second": 2.052,
      "step": 72
    },
    {
      "epoch": 5.52,
      "learning_rate": 0.00014285714285714284,
      "loss": 0.5854,
      "step": 80
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8064516129032258,
      "eval_loss": 0.5508103370666504,
      "eval_runtime": 1.948,
      "eval_samples_per_second": 31.828,
      "eval_steps_per_second": 2.053,
      "step": 87
    },
    {
      "epoch": 6.21,
      "learning_rate": 0.00013988095238095236,
      "loss": 0.5151,
      "step": 90
    },
    {
      "epoch": 6.9,
      "learning_rate": 0.00013690476190476189,
      "loss": 0.4581,
      "step": 100
    },
    {
      "epoch": 6.97,
      "eval_accuracy": 0.8548387096774194,
      "eval_loss": 0.3123740255832672,
      "eval_runtime": 1.9775,
      "eval_samples_per_second": 31.353,
      "eval_steps_per_second": 2.023,
      "step": 101
    },
    {
      "epoch": 7.59,
      "learning_rate": 0.00013392857142857144,
      "loss": 0.386,
      "step": 110
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8548387096774194,
      "eval_loss": 0.31687411665916443,
      "eval_runtime": 2.0035,
      "eval_samples_per_second": 30.946,
      "eval_steps_per_second": 1.997,
      "step": 116
    },
    {
      "epoch": 8.28,
      "learning_rate": 0.00013095238095238093,
      "loss": 0.4182,
      "step": 120
    },
    {
      "epoch": 8.97,
      "learning_rate": 0.00012797619047619045,
      "loss": 0.347,
      "step": 130
    },
    {
      "epoch": 8.97,
      "eval_accuracy": 0.9193548387096774,
      "eval_loss": 0.22072885930538177,
      "eval_runtime": 2.034,
      "eval_samples_per_second": 30.482,
      "eval_steps_per_second": 1.967,
      "step": 130
    },
    {
      "epoch": 9.66,
      "learning_rate": 0.000125,
      "loss": 0.3873,
      "step": 140
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.8225806451612904,
      "eval_loss": 0.5968738198280334,
      "eval_runtime": 1.986,
      "eval_samples_per_second": 31.219,
      "eval_steps_per_second": 2.014,
      "step": 145
    },
    {
      "epoch": 10.34,
      "learning_rate": 0.00012202380952380951,
      "loss": 0.3508,
      "step": 150
    },
    {
      "epoch": 10.97,
      "eval_accuracy": 0.8870967741935484,
      "eval_loss": 0.342462956905365,
      "eval_runtime": 1.981,
      "eval_samples_per_second": 31.298,
      "eval_steps_per_second": 2.019,
      "step": 159
    },
    {
      "epoch": 11.03,
      "learning_rate": 0.00011904761904761903,
      "loss": 0.3437,
      "step": 160
    },
    {
      "epoch": 11.72,
      "learning_rate": 0.00011607142857142857,
      "loss": 0.274,
      "step": 170
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.8709677419354839,
      "eval_loss": 0.3376210927963257,
      "eval_runtime": 2.0573,
      "eval_samples_per_second": 30.136,
      "eval_steps_per_second": 1.944,
      "step": 174
    },
    {
      "epoch": 12.41,
      "learning_rate": 0.00011309523809523808,
      "loss": 0.2615,
      "step": 180
    },
    {
      "epoch": 12.97,
      "eval_accuracy": 0.8709677419354839,
      "eval_loss": 0.4912601709365845,
      "eval_runtime": 2.019,
      "eval_samples_per_second": 30.709,
      "eval_steps_per_second": 1.981,
      "step": 188
    },
    {
      "epoch": 13.1,
      "learning_rate": 0.0001101190476190476,
      "loss": 0.2471,
      "step": 190
    },
    {
      "epoch": 13.79,
      "learning_rate": 0.00010714285714285714,
      "loss": 0.3118,
      "step": 200
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.8870967741935484,
      "eval_loss": 0.40341296792030334,
      "eval_runtime": 1.9425,
      "eval_samples_per_second": 31.918,
      "eval_steps_per_second": 2.059,
      "step": 203
    },
    {
      "epoch": 14.48,
      "learning_rate": 0.00010416666666666666,
      "loss": 0.2205,
      "step": 210
    },
    {
      "epoch": 14.97,
      "eval_accuracy": 0.8709677419354839,
      "eval_loss": 0.3167090117931366,
      "eval_runtime": 1.966,
      "eval_samples_per_second": 31.537,
      "eval_steps_per_second": 2.035,
      "step": 217
    },
    {
      "epoch": 15.17,
      "learning_rate": 0.00010119047619047618,
      "loss": 0.2102,
      "step": 220
    },
    {
      "epoch": 15.86,
      "learning_rate": 9.82142857142857e-05,
      "loss": 0.2325,
      "step": 230
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.8870967741935484,
      "eval_loss": 0.3042737543582916,
      "eval_runtime": 1.9585,
      "eval_samples_per_second": 31.658,
      "eval_steps_per_second": 2.042,
      "step": 232
    },
    {
      "epoch": 16.55,
      "learning_rate": 9.523809523809523e-05,
      "loss": 0.1914,
      "step": 240
    },
    {
      "epoch": 16.97,
      "eval_accuracy": 0.8225806451612904,
      "eval_loss": 0.4256087839603424,
      "eval_runtime": 1.9625,
      "eval_samples_per_second": 31.593,
      "eval_steps_per_second": 2.038,
      "step": 246
    },
    {
      "epoch": 17.24,
      "learning_rate": 9.226190476190476e-05,
      "loss": 0.2044,
      "step": 250
    },
    {
      "epoch": 17.93,
      "learning_rate": 8.928571428571427e-05,
      "loss": 0.1997,
      "step": 260
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.8548387096774194,
      "eval_loss": 0.37694820761680603,
      "eval_runtime": 2.009,
      "eval_samples_per_second": 30.862,
      "eval_steps_per_second": 1.991,
      "step": 261
    },
    {
      "epoch": 18.62,
      "learning_rate": 8.63095238095238e-05,
      "loss": 0.1752,
      "step": 270
    },
    {
      "epoch": 18.97,
      "eval_accuracy": 0.8548387096774194,
      "eval_loss": 0.587546169757843,
      "eval_runtime": 2.1045,
      "eval_samples_per_second": 29.461,
      "eval_steps_per_second": 1.901,
      "step": 275
    },
    {
      "epoch": 19.31,
      "learning_rate": 8.333333333333333e-05,
      "loss": 0.237,
      "step": 280
    },
    {
      "epoch": 20.0,
      "learning_rate": 8.035714285714285e-05,
      "loss": 0.1685,
      "step": 290
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.8870967741935484,
      "eval_loss": 0.41043660044670105,
      "eval_runtime": 1.8789,
      "eval_samples_per_second": 32.997,
      "eval_steps_per_second": 2.129,
      "step": 290
    },
    {
      "epoch": 20.69,
      "learning_rate": 7.738095238095239e-05,
      "loss": 0.1736,
      "step": 300
    },
    {
      "epoch": 20.97,
      "eval_accuracy": 0.8548387096774194,
      "eval_loss": 0.5480897426605225,
      "eval_runtime": 1.8865,
      "eval_samples_per_second": 32.866,
      "eval_steps_per_second": 2.12,
      "step": 304
    },
    {
      "epoch": 21.38,
      "learning_rate": 7.44047619047619e-05,
      "loss": 0.1901,
      "step": 310
    },
    {
      "epoch": 22.0,
      "eval_accuracy": 0.9032258064516129,
      "eval_loss": 0.3800370693206787,
      "eval_runtime": 1.916,
      "eval_samples_per_second": 32.36,
      "eval_steps_per_second": 2.088,
      "step": 319
    },
    {
      "epoch": 22.07,
      "learning_rate": 7.142857142857142e-05,
      "loss": 0.1718,
      "step": 320
    },
    {
      "epoch": 22.76,
      "learning_rate": 6.845238095238094e-05,
      "loss": 0.1426,
      "step": 330
    },
    {
      "epoch": 22.97,
      "eval_accuracy": 0.8870967741935484,
      "eval_loss": 0.44246000051498413,
      "eval_runtime": 1.939,
      "eval_samples_per_second": 31.976,
      "eval_steps_per_second": 2.063,
      "step": 333
    },
    {
      "epoch": 23.45,
      "learning_rate": 6.547619047619047e-05,
      "loss": 0.1251,
      "step": 340
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.9032258064516129,
      "eval_loss": 0.3373814523220062,
      "eval_runtime": 1.9735,
      "eval_samples_per_second": 31.416,
      "eval_steps_per_second": 2.027,
      "step": 348
    },
    {
      "epoch": 24.14,
      "learning_rate": 6.25e-05,
      "loss": 0.1329,
      "step": 350
    },
    {
      "epoch": 24.83,
      "learning_rate": 5.952380952380952e-05,
      "loss": 0.1326,
      "step": 360
    },
    {
      "epoch": 24.97,
      "eval_accuracy": 0.8870967741935484,
      "eval_loss": 0.3627336323261261,
      "eval_runtime": 1.951,
      "eval_samples_per_second": 31.779,
      "eval_steps_per_second": 2.05,
      "step": 362
    },
    {
      "epoch": 25.52,
      "learning_rate": 5.654761904761904e-05,
      "loss": 0.1271,
      "step": 370
    },
    {
      "epoch": 26.0,
      "eval_accuracy": 0.8709677419354839,
      "eval_loss": 0.4767535328865051,
      "eval_runtime": 1.977,
      "eval_samples_per_second": 31.361,
      "eval_steps_per_second": 2.023,
      "step": 377
    },
    {
      "epoch": 26.21,
      "learning_rate": 5.357142857142857e-05,
      "loss": 0.1415,
      "step": 380
    },
    {
      "epoch": 26.9,
      "learning_rate": 5.059523809523809e-05,
      "loss": 0.1835,
      "step": 390
    },
    {
      "epoch": 26.97,
      "eval_accuracy": 0.8709677419354839,
      "eval_loss": 0.5603958368301392,
      "eval_runtime": 2.171,
      "eval_samples_per_second": 28.558,
      "eval_steps_per_second": 1.842,
      "step": 391
    },
    {
      "epoch": 27.59,
      "learning_rate": 4.7619047619047614e-05,
      "loss": 0.1378,
      "step": 400
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.8870967741935484,
      "eval_loss": 0.4130818247795105,
      "eval_runtime": 2.0055,
      "eval_samples_per_second": 30.916,
      "eval_steps_per_second": 1.995,
      "step": 406
    },
    {
      "epoch": 28.28,
      "learning_rate": 4.4642857142857136e-05,
      "loss": 0.1253,
      "step": 410
    },
    {
      "epoch": 28.97,
      "learning_rate": 4.1666666666666665e-05,
      "loss": 0.1349,
      "step": 420
    },
    {
      "epoch": 28.97,
      "eval_accuracy": 0.8548387096774194,
      "eval_loss": 0.5103474259376526,
      "eval_runtime": 1.9155,
      "eval_samples_per_second": 32.368,
      "eval_steps_per_second": 2.088,
      "step": 420
    },
    {
      "epoch": 29.66,
      "learning_rate": 3.8690476190476195e-05,
      "loss": 0.0999,
      "step": 430
    },
    {
      "epoch": 30.0,
      "eval_accuracy": 0.9193548387096774,
      "eval_loss": 0.37231481075286865,
      "eval_runtime": 1.94,
      "eval_samples_per_second": 31.959,
      "eval_steps_per_second": 2.062,
      "step": 435
    },
    {
      "epoch": 30.34,
      "learning_rate": 3.571428571428571e-05,
      "loss": 0.1198,
      "step": 440
    },
    {
      "epoch": 30.97,
      "eval_accuracy": 0.8709677419354839,
      "eval_loss": 0.5360597968101501,
      "eval_runtime": 1.9235,
      "eval_samples_per_second": 32.233,
      "eval_steps_per_second": 2.08,
      "step": 449
    },
    {
      "epoch": 31.03,
      "learning_rate": 3.273809523809523e-05,
      "loss": 0.1301,
      "step": 450
    },
    {
      "epoch": 31.72,
      "learning_rate": 2.976190476190476e-05,
      "loss": 0.1195,
      "step": 460
    },
    {
      "epoch": 32.0,
      "eval_accuracy": 0.8870967741935484,
      "eval_loss": 0.41935569047927856,
      "eval_runtime": 1.9165,
      "eval_samples_per_second": 32.351,
      "eval_steps_per_second": 2.087,
      "step": 464
    },
    {
      "epoch": 32.41,
      "learning_rate": 2.6785714285714284e-05,
      "loss": 0.0766,
      "step": 470
    },
    {
      "epoch": 32.97,
      "eval_accuracy": 0.8870967741935484,
      "eval_loss": 0.4133478105068207,
      "eval_runtime": 1.924,
      "eval_samples_per_second": 32.225,
      "eval_steps_per_second": 2.079,
      "step": 478
    },
    {
      "epoch": 33.1,
      "learning_rate": 2.3809523809523807e-05,
      "loss": 0.1043,
      "step": 480
    },
    {
      "epoch": 33.79,
      "learning_rate": 2.0833333333333333e-05,
      "loss": 0.0862,
      "step": 490
    },
    {
      "epoch": 34.0,
      "eval_accuracy": 0.9032258064516129,
      "eval_loss": 0.42390120029449463,
      "eval_runtime": 1.9,
      "eval_samples_per_second": 32.632,
      "eval_steps_per_second": 2.105,
      "step": 493
    },
    {
      "epoch": 34.48,
      "learning_rate": 1.7857142857142855e-05,
      "loss": 0.1048,
      "step": 500
    },
    {
      "epoch": 34.97,
      "eval_accuracy": 0.9193548387096774,
      "eval_loss": 0.4120253920555115,
      "eval_runtime": 1.954,
      "eval_samples_per_second": 31.73,
      "eval_steps_per_second": 2.047,
      "step": 507
    },
    {
      "epoch": 35.17,
      "learning_rate": 1.488095238095238e-05,
      "loss": 0.0884,
      "step": 510
    },
    {
      "epoch": 35.86,
      "learning_rate": 1.1904761904761903e-05,
      "loss": 0.0902,
      "step": 520
    },
    {
      "epoch": 36.0,
      "eval_accuracy": 0.9032258064516129,
      "eval_loss": 0.44083285331726074,
      "eval_runtime": 1.9034,
      "eval_samples_per_second": 32.573,
      "eval_steps_per_second": 2.101,
      "step": 522
    },
    {
      "epoch": 36.55,
      "learning_rate": 8.928571428571428e-06,
      "loss": 0.088,
      "step": 530
    },
    {
      "epoch": 36.97,
      "eval_accuracy": 0.9032258064516129,
      "eval_loss": 0.4435848295688629,
      "eval_runtime": 1.8659,
      "eval_samples_per_second": 33.227,
      "eval_steps_per_second": 2.144,
      "step": 536
    },
    {
      "epoch": 37.24,
      "learning_rate": 5.952380952380952e-06,
      "loss": 0.0864,
      "step": 540
    },
    {
      "epoch": 37.93,
      "learning_rate": 2.976190476190476e-06,
      "loss": 0.089,
      "step": 550
    },
    {
      "epoch": 38.0,
      "eval_accuracy": 0.9032258064516129,
      "eval_loss": 0.46484920382499695,
      "eval_runtime": 1.9695,
      "eval_samples_per_second": 31.481,
      "eval_steps_per_second": 2.031,
      "step": 551
    },
    {
      "epoch": 38.62,
      "learning_rate": 0.0,
      "loss": 0.1089,
      "step": 560
    },
    {
      "epoch": 38.62,
      "eval_accuracy": 0.8870967741935484,
      "eval_loss": 0.46501168608665466,
      "eval_runtime": 2.034,
      "eval_samples_per_second": 30.482,
      "eval_steps_per_second": 1.967,
      "step": 560
    },
    {
      "epoch": 38.62,
      "step": 560,
      "total_flos": 1.1660953582043136e+18,
      "train_loss": 0.30870234380875317,
      "train_runtime": 1509.6082,
      "train_samples_per_second": 24.589,
      "train_steps_per_second": 0.371
    }
  ],
  "logging_steps": 10,
  "max_steps": 560,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 40,
  "save_steps": 500,
  "total_flos": 1.1660953582043136e+18,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}