{
  "best_metric": 0.7139084935188293,
  "best_model_checkpoint": "tiny_bert_km_50_v1_wnli/checkpoint-6",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 21,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.017745852470398,
      "learning_rate": 4.9e-05,
      "loss": 0.7038,
      "step": 3
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.38028169014084506,
      "eval_loss": 0.7231513857841492,
      "eval_runtime": 0.0396,
      "eval_samples_per_second": 1790.713,
      "eval_steps_per_second": 25.221,
      "step": 3
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6744726300239563,
      "learning_rate": 4.8e-05,
      "loss": 0.6955,
      "step": 6
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.38028169014084506,
      "eval_loss": 0.7139084935188293,
      "eval_runtime": 0.0395,
      "eval_samples_per_second": 1799.402,
      "eval_steps_per_second": 25.344,
      "step": 6
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.6699543595314026,
      "learning_rate": 4.7e-05,
      "loss": 0.692,
      "step": 9
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.36619718309859156,
      "eval_loss": 0.7201256155967712,
      "eval_runtime": 0.0472,
      "eval_samples_per_second": 1504.406,
      "eval_steps_per_second": 21.189,
      "step": 9
    },
    {
      "epoch": 4.0,
      "grad_norm": 0.7492492198944092,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.6926,
      "step": 12
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.352112676056338,
      "eval_loss": 0.7315689921379089,
      "eval_runtime": 0.0383,
      "eval_samples_per_second": 1855.366,
      "eval_steps_per_second": 26.132,
      "step": 12
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.8681108951568604,
      "learning_rate": 4.5e-05,
      "loss": 0.689,
      "step": 15
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.28169014084507044,
      "eval_loss": 0.7487897872924805,
      "eval_runtime": 0.0375,
      "eval_samples_per_second": 1894.591,
      "eval_steps_per_second": 26.684,
      "step": 15
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.8175243735313416,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.6869,
      "step": 18
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.28169014084507044,
      "eval_loss": 0.7664502263069153,
      "eval_runtime": 0.039,
      "eval_samples_per_second": 1820.511,
      "eval_steps_per_second": 25.641,
      "step": 18
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.7496082186698914,
      "learning_rate": 4.3e-05,
      "loss": 0.6912,
      "step": 21
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.23943661971830985,
      "eval_loss": 0.771676778793335,
      "eval_runtime": 0.0377,
      "eval_samples_per_second": 1880.937,
      "eval_steps_per_second": 26.492,
      "step": 21
    },
    {
      "epoch": 7.0,
      "step": 21,
      "total_flos": 116563804707840.0,
      "train_loss": 0.6929929029373896,
      "train_runtime": 8.4984,
      "train_samples_per_second": 3735.985,
      "train_steps_per_second": 17.65
    }
  ],
  "logging_steps": 1,
  "max_steps": 150,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "total_flos": 116563804707840.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}