trainer_state.json

{
  "best_metric": 0.36275917291641235,
  "best_model_checkpoint": "bert_uncased_L-4_H-256_A-4_sst2/checkpoint-528",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 1848,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 8.933422088623047,
      "learning_rate": 4.9e-05,
      "loss": 0.4143,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8130733944954128,
      "eval_loss": 0.41190946102142334,
      "eval_runtime": 0.3091,
      "eval_samples_per_second": 2821.207,
      "eval_steps_per_second": 12.941,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.186891555786133,
      "learning_rate": 4.8e-05,
      "loss": 0.2672,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8428899082568807,
      "eval_loss": 0.36275917291641235,
      "eval_runtime": 0.3107,
      "eval_samples_per_second": 2806.986,
      "eval_steps_per_second": 12.876,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.041283130645752,
      "learning_rate": 4.7e-05,
      "loss": 0.2117,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8497706422018348,
      "eval_loss": 0.3785320520401001,
      "eval_runtime": 0.3204,
      "eval_samples_per_second": 2721.843,
      "eval_steps_per_second": 12.486,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.730945587158203,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1736,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8474770642201835,
      "eval_loss": 0.41394656896591187,
      "eval_runtime": 0.3097,
      "eval_samples_per_second": 2815.921,
      "eval_steps_per_second": 12.917,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 8.553589820861816,
      "learning_rate": 4.5e-05,
      "loss": 0.1493,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8532110091743119,
      "eval_loss": 0.42312726378440857,
      "eval_runtime": 0.32,
      "eval_samples_per_second": 2724.974,
      "eval_steps_per_second": 12.5,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 7.31311559677124,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.1294,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8543577981651376,
      "eval_loss": 0.44111400842666626,
      "eval_runtime": 0.3158,
      "eval_samples_per_second": 2760.954,
      "eval_steps_per_second": 12.665,
      "step": 1584
    },
    {
      "epoch": 7.0,
      "grad_norm": 9.353928565979004,
      "learning_rate": 4.3e-05,
      "loss": 0.1159,
      "step": 1848
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8589449541284404,
      "eval_loss": 0.4574843645095825,
      "eval_runtime": 0.3048,
      "eval_samples_per_second": 2861.12,
      "eval_steps_per_second": 13.124,
      "step": 1848
    },
    {
      "epoch": 7.0,
      "step": 1848,
      "total_flos": 2335961353872384.0,
      "train_loss": 0.20875395428050647,
      "train_runtime": 224.6839,
      "train_samples_per_second": 14987.5,
      "train_steps_per_second": 58.749
    }
  ],
  "logging_steps": 1,
  "max_steps": 13200,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 50,
  "save_steps": 500,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 5,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 5
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2335961353872384.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}