{
  "best_metric": 0.4193161725997925,
  "best_model_checkpoint": "tiny_bert_km_50_v1_qqp/checkpoint-4266",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 11376,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.748605489730835,
      "learning_rate": 4.9e-05,
      "loss": 0.5011,
      "step": 1422
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7840217660153351,
      "eval_combined_score": 0.732481505632631,
      "eval_f1": 0.6809412452499269,
      "eval_loss": 0.4513613283634186,
      "eval_runtime": 15.0073,
      "eval_samples_per_second": 2694.026,
      "eval_steps_per_second": 10.528,
      "step": 1422
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.0197439193725586,
      "learning_rate": 4.8e-05,
      "loss": 0.4146,
      "step": 2844
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7996784565916398,
      "eval_combined_score": 0.7551677267417489,
      "eval_f1": 0.7106569968918581,
      "eval_loss": 0.42334410548210144,
      "eval_runtime": 14.0744,
      "eval_samples_per_second": 2872.589,
      "eval_steps_per_second": 11.226,
      "step": 2844
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.5988855361938477,
      "learning_rate": 4.7e-05,
      "loss": 0.3605,
      "step": 4266
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8108582735592382,
      "eval_combined_score": 0.7698776779218888,
      "eval_f1": 0.7288970822845393,
      "eval_loss": 0.4193161725997925,
      "eval_runtime": 14.0194,
      "eval_samples_per_second": 2883.855,
      "eval_steps_per_second": 11.27,
      "step": 4266
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.5036709308624268,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.3153,
      "step": 5688
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8142715805095226,
      "eval_combined_score": 0.7768661951678455,
      "eval_f1": 0.7394608098261685,
      "eval_loss": 0.42982715368270874,
      "eval_runtime": 14.0649,
      "eval_samples_per_second": 2874.526,
      "eval_steps_per_second": 11.234,
      "step": 5688
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.8022961616516113,
      "learning_rate": 4.5e-05,
      "loss": 0.2768,
      "step": 7110
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8161019045263418,
      "eval_combined_score": 0.7811781492871632,
      "eval_f1": 0.7462543940479847,
      "eval_loss": 0.4367225170135498,
      "eval_runtime": 14.2004,
      "eval_samples_per_second": 2847.095,
      "eval_steps_per_second": 11.126,
      "step": 7110
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.0807063579559326,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.245,
      "step": 8532
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8136779619094732,
      "eval_combined_score": 0.7862144813550569,
      "eval_f1": 0.7587510008006405,
      "eval_loss": 0.4542311131954193,
      "eval_runtime": 14.0405,
      "eval_samples_per_second": 2879.525,
      "eval_steps_per_second": 11.253,
      "step": 8532
    },
    {
      "epoch": 7.0,
      "grad_norm": 4.5525922775268555,
      "learning_rate": 4.3e-05,
      "loss": 0.2174,
      "step": 9954
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8195152114766263,
      "eval_combined_score": 0.7896423794008647,
      "eval_f1": 0.7597695473251029,
      "eval_loss": 0.47380971908569336,
      "eval_runtime": 14.0747,
      "eval_samples_per_second": 2872.521,
      "eval_steps_per_second": 11.226,
      "step": 9954
    },
    {
      "epoch": 8.0,
      "grad_norm": 3.2192418575286865,
      "learning_rate": 4.2e-05,
      "loss": 0.1927,
      "step": 11376
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.823695275785308,
      "eval_combined_score": 0.7905984033549933,
      "eval_f1": 0.7575015309246785,
      "eval_loss": 0.5163273811340332,
      "eval_runtime": 14.1681,
      "eval_samples_per_second": 2853.585,
      "eval_steps_per_second": 11.152,
      "step": 11376
    },
    {
      "epoch": 8.0,
      "step": 11376,
      "total_flos": 7.633075201391002e+16,
      "train_loss": 0.315434974960134,
      "train_runtime": 2030.1953,
      "train_samples_per_second": 8960.862,
      "train_steps_per_second": 35.021
    }
  ],
  "logging_steps": 1,
  "max_steps": 71100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "total_flos": 7.633075201391002e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}