run-6/checkpoint-33676/trainer_state.json

{
  "best_metric": 0.9151376146788991,
  "best_model_checkpoint": "Tech-oriented/best_model_bert_uncasedbert-base-uncased-finetuned-sst2/run-6/checkpoint-33676",
  "epoch": 4.0,
  "eval_steps": 500,
  "global_step": 33676,
  "is_hyper_param_search": true,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5,
      "grad_norm": 35.4764289855957,
      "learning_rate": 1.4625672140464685e-05,
      "loss": 0.452,
      "step": 4209
    },
    {
      "epoch": 1.0,
      "grad_norm": 2.615936040878296,
      "learning_rate": 1.2536574029383954e-05,
      "loss": 0.354,
      "step": 8418
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9139908256880734,
      "eval_loss": 0.36559009552001953,
      "eval_runtime": 2.4317,
      "eval_samples_per_second": 358.596,
      "eval_steps_per_second": 22.618,
      "step": 8419
    },
    {
      "epoch": 1.5,
      "grad_norm": 1.231093406677246,
      "learning_rate": 1.0447475918303226e-05,
      "loss": 0.3236,
      "step": 12627
    },
    {
      "epoch": 2.0,
      "grad_norm": 66.64139556884766,
      "learning_rate": 8.358377807222495e-06,
      "loss": 0.303,
      "step": 16836
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.9025229357798165,
      "eval_loss": 0.4449034631252289,
      "eval_runtime": 2.4314,
      "eval_samples_per_second": 358.637,
      "eval_steps_per_second": 22.62,
      "step": 16838
    },
    {
      "epoch": 2.5,
      "grad_norm": 15.909801483154297,
      "learning_rate": 6.269279696141766e-06,
      "loss": 0.2693,
      "step": 21045
    },
    {
      "epoch": 3.0,
      "grad_norm": 3.272183895111084,
      "learning_rate": 4.180181585061036e-06,
      "loss": 0.2588,
      "step": 25254
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9105504587155964,
      "eval_loss": 0.4759577810764313,
      "eval_runtime": 2.4404,
      "eval_samples_per_second": 357.326,
      "eval_steps_per_second": 22.538,
      "step": 25257
    },
    {
      "epoch": 3.5,
      "grad_norm": 21.020828247070312,
      "learning_rate": 2.0910834739803075e-06,
      "loss": 0.2357,
      "step": 29463
    },
    {
      "epoch": 4.0,
      "grad_norm": 8.535745620727539,
      "learning_rate": 1.98536289957779e-09,
      "loss": 0.232,
      "step": 33672
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9151376146788991,
      "eval_loss": 0.4764806628227234,
      "eval_runtime": 2.4265,
      "eval_samples_per_second": 359.361,
      "eval_steps_per_second": 22.666,
      "step": 33676
    }
  ],
  "logging_steps": 4209,
  "max_steps": 33676,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "total_flos": 5500248223430640.0,
  "train_batch_size": 8,
  "trial_name": null,
  "trial_params": {
    "learning_rate": 1.6714770251545413e-05,
    "num_train_epochs": 4,
    "per_device_train_batch_size": 8,
    "seed": 32,
    "weight_decay": 0.41710891288145024
  }
}