{
  "best_metric": 0.6304347826086957,
  "best_model_checkpoint": "SW2-DMAE-DA\\checkpoint-368",
  "epoch": 38.26086956521739,
  "eval_steps": 500,
  "global_step": 440,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.87,
      "learning_rate": 9.090909090909091e-06,
      "loss": 6.464,
      "step": 10
    },
    {
      "epoch": 0.96,
      "eval_accuracy": 0.10869565217391304,
      "eval_loss": 7.919026851654053,
      "eval_runtime": 0.5926,
      "eval_samples_per_second": 77.618,
      "eval_steps_per_second": 5.062,
      "step": 11
    },
    {
      "epoch": 1.74,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 6.5496,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.10869565217391304,
      "eval_loss": 7.5824174880981445,
      "eval_runtime": 0.6062,
      "eval_samples_per_second": 75.889,
      "eval_steps_per_second": 4.949,
      "step": 23
    },
    {
      "epoch": 2.61,
      "learning_rate": 2.7272727272727273e-05,
      "loss": 6.1541,
      "step": 30
    },
    {
      "epoch": 2.96,
      "eval_accuracy": 0.10869565217391304,
      "eval_loss": 6.415637016296387,
      "eval_runtime": 0.6292,
      "eval_samples_per_second": 73.114,
      "eval_steps_per_second": 4.768,
      "step": 34
    },
    {
      "epoch": 3.48,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 5.2649,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.10869565217391304,
      "eval_loss": 4.506710052490234,
      "eval_runtime": 0.6096,
      "eval_samples_per_second": 75.454,
      "eval_steps_per_second": 4.921,
      "step": 46
    },
    {
      "epoch": 4.35,
      "learning_rate": 3.93939393939394e-05,
      "loss": 4.1175,
      "step": 50
    },
    {
      "epoch": 4.96,
      "eval_accuracy": 0.10869565217391304,
      "eval_loss": 2.724062442779541,
      "eval_runtime": 0.6126,
      "eval_samples_per_second": 75.084,
      "eval_steps_per_second": 4.897,
      "step": 57
    },
    {
      "epoch": 5.22,
      "learning_rate": 3.838383838383839e-05,
      "loss": 2.8424,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.10869565217391304,
      "eval_loss": 1.562261939048767,
      "eval_runtime": 0.6031,
      "eval_samples_per_second": 76.268,
      "eval_steps_per_second": 4.974,
      "step": 69
    },
    {
      "epoch": 6.09,
      "learning_rate": 3.7373737373737376e-05,
      "loss": 1.8484,
      "step": 70
    },
    {
      "epoch": 6.96,
      "learning_rate": 3.6363636363636364e-05,
      "loss": 1.4376,
      "step": 80
    },
    {
      "epoch": 6.96,
      "eval_accuracy": 0.10869565217391304,
      "eval_loss": 1.40028977394104,
      "eval_runtime": 0.5771,
      "eval_samples_per_second": 79.703,
      "eval_steps_per_second": 5.198,
      "step": 80
    },
    {
      "epoch": 7.83,
      "learning_rate": 3.535353535353536e-05,
      "loss": 1.4054,
      "step": 90
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.10869565217391304,
      "eval_loss": 1.4054746627807617,
      "eval_runtime": 0.6232,
      "eval_samples_per_second": 73.818,
      "eval_steps_per_second": 4.814,
      "step": 92
    },
    {
      "epoch": 8.7,
      "learning_rate": 3.434343434343435e-05,
      "loss": 1.3798,
      "step": 100
    },
    {
      "epoch": 8.96,
      "eval_accuracy": 0.45652173913043476,
      "eval_loss": 1.3465980291366577,
      "eval_runtime": 0.6126,
      "eval_samples_per_second": 75.084,
      "eval_steps_per_second": 4.897,
      "step": 103
    },
    {
      "epoch": 9.57,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 1.3331,
      "step": 110
    },
    {
      "epoch": 10.0,
      "eval_accuracy": 0.15217391304347827,
      "eval_loss": 1.4385319948196411,
      "eval_runtime": 0.5961,
      "eval_samples_per_second": 77.164,
      "eval_steps_per_second": 5.032,
      "step": 115
    },
    {
      "epoch": 10.43,
      "learning_rate": 3.232323232323232e-05,
      "loss": 1.2736,
      "step": 120
    },
    {
      "epoch": 10.96,
      "eval_accuracy": 0.2391304347826087,
      "eval_loss": 1.3105802536010742,
      "eval_runtime": 0.5992,
      "eval_samples_per_second": 76.774,
      "eval_steps_per_second": 5.007,
      "step": 126
    },
    {
      "epoch": 11.3,
      "learning_rate": 3.131313131313132e-05,
      "loss": 1.2127,
      "step": 130
    },
    {
      "epoch": 12.0,
      "eval_accuracy": 0.1956521739130435,
      "eval_loss": 1.2908252477645874,
      "eval_runtime": 0.6151,
      "eval_samples_per_second": 74.78,
      "eval_steps_per_second": 4.877,
      "step": 138
    },
    {
      "epoch": 12.17,
      "learning_rate": 3.0303030303030306e-05,
      "loss": 1.2531,
      "step": 140
    },
    {
      "epoch": 12.96,
      "eval_accuracy": 0.5,
      "eval_loss": 1.2544833421707153,
      "eval_runtime": 0.5956,
      "eval_samples_per_second": 77.228,
      "eval_steps_per_second": 5.037,
      "step": 149
    },
    {
      "epoch": 13.04,
      "learning_rate": 2.9292929292929297e-05,
      "loss": 1.2027,
      "step": 150
    },
    {
      "epoch": 13.91,
      "learning_rate": 2.8282828282828285e-05,
      "loss": 1.0972,
      "step": 160
    },
    {
      "epoch": 14.0,
      "eval_accuracy": 0.34782608695652173,
      "eval_loss": 1.25150465965271,
      "eval_runtime": 0.6331,
      "eval_samples_per_second": 72.654,
      "eval_steps_per_second": 4.738,
      "step": 161
    },
    {
      "epoch": 14.78,
      "learning_rate": 2.7272727272727273e-05,
      "loss": 1.0029,
      "step": 170
    },
    {
      "epoch": 14.96,
      "eval_accuracy": 0.2608695652173913,
      "eval_loss": 1.2238019704818726,
      "eval_runtime": 0.6177,
      "eval_samples_per_second": 74.474,
      "eval_steps_per_second": 4.857,
      "step": 172
    },
    {
      "epoch": 15.65,
      "learning_rate": 2.6262626262626265e-05,
      "loss": 1.0141,
      "step": 180
    },
    {
      "epoch": 16.0,
      "eval_accuracy": 0.3695652173913043,
      "eval_loss": 1.20670485496521,
      "eval_runtime": 0.5971,
      "eval_samples_per_second": 77.033,
      "eval_steps_per_second": 5.024,
      "step": 184
    },
    {
      "epoch": 16.52,
      "learning_rate": 2.5252525252525253e-05,
      "loss": 0.9129,
      "step": 190
    },
    {
      "epoch": 16.96,
      "eval_accuracy": 0.5652173913043478,
      "eval_loss": 1.1149216890335083,
      "eval_runtime": 0.6201,
      "eval_samples_per_second": 74.177,
      "eval_steps_per_second": 4.838,
      "step": 195
    },
    {
      "epoch": 17.39,
      "learning_rate": 2.4242424242424244e-05,
      "loss": 0.9157,
      "step": 200
    },
    {
      "epoch": 18.0,
      "eval_accuracy": 0.391304347826087,
      "eval_loss": 1.1956795454025269,
      "eval_runtime": 0.6221,
      "eval_samples_per_second": 73.938,
      "eval_steps_per_second": 4.822,
      "step": 207
    },
    {
      "epoch": 18.26,
      "learning_rate": 2.3232323232323232e-05,
      "loss": 0.8516,
      "step": 210
    },
    {
      "epoch": 18.96,
      "eval_accuracy": 0.5434782608695652,
      "eval_loss": 1.0033633708953857,
      "eval_runtime": 0.5911,
      "eval_samples_per_second": 77.815,
      "eval_steps_per_second": 5.075,
      "step": 218
    },
    {
      "epoch": 19.13,
      "learning_rate": 2.2222222222222227e-05,
      "loss": 0.8139,
      "step": 220
    },
    {
      "epoch": 20.0,
      "learning_rate": 2.121212121212121e-05,
      "loss": 0.7804,
      "step": 230
    },
    {
      "epoch": 20.0,
      "eval_accuracy": 0.4782608695652174,
      "eval_loss": 0.9990861415863037,
      "eval_runtime": 0.6019,
      "eval_samples_per_second": 76.429,
      "eval_steps_per_second": 4.984,
      "step": 230
    },
    {
      "epoch": 20.87,
      "learning_rate": 2.0202020202020206e-05,
      "loss": 0.7328,
      "step": 240
    },
    {
      "epoch": 20.96,
      "eval_accuracy": 0.5869565217391305,
      "eval_loss": 0.9839891195297241,
      "eval_runtime": 0.6041,
      "eval_samples_per_second": 76.142,
      "eval_steps_per_second": 4.966,
      "step": 241
    },
    {
      "epoch": 21.74,
      "learning_rate": 1.9191919191919194e-05,
      "loss": 0.7101,
      "step": 250
    },
    {
      "epoch": 22.0,
      "eval_accuracy": 0.5434782608695652,
      "eval_loss": 0.9661365747451782,
      "eval_runtime": 0.6302,
      "eval_samples_per_second": 72.989,
      "eval_steps_per_second": 4.76,
      "step": 253
    },
    {
      "epoch": 22.61,
      "learning_rate": 1.8181818181818182e-05,
      "loss": 0.7099,
      "step": 260
    },
    {
      "epoch": 22.96,
      "eval_accuracy": 0.5434782608695652,
      "eval_loss": 0.9392473697662354,
      "eval_runtime": 0.5961,
      "eval_samples_per_second": 77.164,
      "eval_steps_per_second": 5.032,
      "step": 264
    },
    {
      "epoch": 23.48,
      "learning_rate": 1.7171717171717173e-05,
      "loss": 0.7238,
      "step": 270
    },
    {
      "epoch": 24.0,
      "eval_accuracy": 0.5,
      "eval_loss": 0.9552508592605591,
      "eval_runtime": 0.6066,
      "eval_samples_per_second": 75.827,
      "eval_steps_per_second": 4.945,
      "step": 276
    },
    {
      "epoch": 24.35,
      "learning_rate": 1.616161616161616e-05,
      "loss": 0.6605,
      "step": 280
    },
    {
      "epoch": 24.96,
      "eval_accuracy": 0.5434782608695652,
      "eval_loss": 0.9571367502212524,
      "eval_runtime": 0.5891,
      "eval_samples_per_second": 78.081,
      "eval_steps_per_second": 5.092,
      "step": 287
    },
    {
      "epoch": 25.22,
      "learning_rate": 1.5151515151515153e-05,
      "loss": 0.639,
      "step": 290
    },
    {
      "epoch": 26.0,
      "eval_accuracy": 0.5652173913043478,
      "eval_loss": 1.0534369945526123,
      "eval_runtime": 0.6126,
      "eval_samples_per_second": 75.084,
      "eval_steps_per_second": 4.897,
      "step": 299
    },
    {
      "epoch": 26.09,
      "learning_rate": 1.4141414141414143e-05,
      "loss": 0.6241,
      "step": 300
    },
    {
      "epoch": 26.96,
      "learning_rate": 1.3131313131313132e-05,
      "loss": 0.6123,
      "step": 310
    },
    {
      "epoch": 26.96,
      "eval_accuracy": 0.6086956521739131,
      "eval_loss": 0.9152448773384094,
      "eval_runtime": 0.7392,
      "eval_samples_per_second": 62.232,
      "eval_steps_per_second": 4.059,
      "step": 310
    },
    {
      "epoch": 27.83,
      "learning_rate": 1.2121212121212122e-05,
      "loss": 0.6021,
      "step": 320
    },
    {
      "epoch": 28.0,
      "eval_accuracy": 0.5869565217391305,
      "eval_loss": 0.8703503608703613,
      "eval_runtime": 0.6151,
      "eval_samples_per_second": 74.78,
      "eval_steps_per_second": 4.877,
      "step": 322
    },
    {
      "epoch": 28.7,
      "learning_rate": 1.1111111111111113e-05,
      "loss": 0.5971,
      "step": 330
    },
    {
      "epoch": 28.96,
      "eval_accuracy": 0.5652173913043478,
      "eval_loss": 0.8726406097412109,
      "eval_runtime": 0.6187,
      "eval_samples_per_second": 74.355,
      "eval_steps_per_second": 4.849,
      "step": 333
    },
    {
      "epoch": 29.57,
      "learning_rate": 1.0101010101010103e-05,
      "loss": 0.5413,
      "step": 340
    },
    {
      "epoch": 30.0,
      "eval_accuracy": 0.5869565217391305,
      "eval_loss": 0.8287004232406616,
      "eval_runtime": 0.5996,
      "eval_samples_per_second": 76.712,
      "eval_steps_per_second": 5.003,
      "step": 345
    },
    {
      "epoch": 30.43,
      "learning_rate": 9.090909090909091e-06,
      "loss": 0.5663,
      "step": 350
    },
    {
      "epoch": 30.96,
      "eval_accuracy": 0.5434782608695652,
      "eval_loss": 0.9271087646484375,
      "eval_runtime": 0.6056,
      "eval_samples_per_second": 75.952,
      "eval_steps_per_second": 4.953,
      "step": 356
    },
    {
      "epoch": 31.3,
      "learning_rate": 8.08080808080808e-06,
      "loss": 0.5343,
      "step": 360
    },
    {
      "epoch": 32.0,
      "eval_accuracy": 0.6304347826086957,
      "eval_loss": 0.8855524659156799,
      "eval_runtime": 0.5976,
      "eval_samples_per_second": 76.969,
      "eval_steps_per_second": 5.02,
      "step": 368
    },
    {
      "epoch": 32.17,
      "learning_rate": 7.070707070707071e-06,
      "loss": 0.525,
      "step": 370
    },
    {
      "epoch": 32.96,
      "eval_accuracy": 0.6086956521739131,
      "eval_loss": 0.8579288721084595,
      "eval_runtime": 0.6126,
      "eval_samples_per_second": 75.084,
      "eval_steps_per_second": 4.897,
      "step": 379
    },
    {
      "epoch": 33.04,
      "learning_rate": 6.060606060606061e-06,
      "loss": 0.5172,
      "step": 380
    },
    {
      "epoch": 33.91,
      "learning_rate": 5.0505050505050515e-06,
      "loss": 0.5447,
      "step": 390
    },
    {
      "epoch": 34.0,
      "eval_accuracy": 0.5869565217391305,
      "eval_loss": 0.8746341466903687,
      "eval_runtime": 0.6122,
      "eval_samples_per_second": 75.144,
      "eval_steps_per_second": 4.901,
      "step": 391
    },
    {
      "epoch": 34.78,
      "learning_rate": 4.04040404040404e-06,
      "loss": 0.5036,
      "step": 400
    },
    {
      "epoch": 34.96,
      "eval_accuracy": 0.5652173913043478,
      "eval_loss": 0.8684332370758057,
      "eval_runtime": 0.6021,
      "eval_samples_per_second": 76.395,
      "eval_steps_per_second": 4.982,
      "step": 402
    },
    {
      "epoch": 35.65,
      "learning_rate": 3.0303030303030305e-06,
      "loss": 0.4918,
      "step": 410
    },
    {
      "epoch": 36.0,
      "eval_accuracy": 0.5869565217391305,
      "eval_loss": 0.8268190622329712,
      "eval_runtime": 0.6312,
      "eval_samples_per_second": 72.881,
      "eval_steps_per_second": 4.753,
      "step": 414
    },
    {
      "epoch": 36.52,
      "learning_rate": 2.02020202020202e-06,
      "loss": 0.503,
      "step": 420
    },
    {
      "epoch": 36.96,
      "eval_accuracy": 0.5869565217391305,
      "eval_loss": 0.8373963832855225,
      "eval_runtime": 0.5726,
      "eval_samples_per_second": 80.33,
      "eval_steps_per_second": 5.239,
      "step": 425
    },
    {
      "epoch": 37.39,
      "learning_rate": 1.01010101010101e-06,
      "loss": 0.5114,
      "step": 430
    },
    {
      "epoch": 38.0,
      "eval_accuracy": 0.6086956521739131,
      "eval_loss": 0.8380029797554016,
      "eval_runtime": 0.6106,
      "eval_samples_per_second": 75.33,
      "eval_steps_per_second": 4.913,
      "step": 437
    },
    {
      "epoch": 38.26,
      "learning_rate": 0.0,
      "loss": 0.5272,
      "step": 440
    },
    {
      "epoch": 38.26,
      "eval_accuracy": 0.6086956521739131,
      "eval_loss": 0.8386858105659485,
      "eval_runtime": 0.6126,
      "eval_samples_per_second": 75.084,
      "eval_steps_per_second": 4.897,
      "step": 440
    },
    {
      "epoch": 38.26,
      "step": 440,
      "total_flos": 8.989085534729011e+17,
      "train_loss": 1.443298595601862,
      "train_runtime": 559.502,
      "train_samples_per_second": 51.617,
      "train_steps_per_second": 0.786
    }
  ],
  "logging_steps": 10,
  "max_steps": 440,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 40,
  "save_steps": 500,
  "total_flos": 8.989085534729011e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}