trainer_state.json

{
  "best_global_step": 10,
  "best_metric": 0.7063684463500977,
  "best_model_checkpoint": "distilbert_km_10_v2_rte/checkpoint-10",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 60,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.3346072435379028,
      "learning_rate": 4.9e-05,
      "loss": 0.7038,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5126353790613718,
      "eval_loss": 0.7063684463500977,
      "eval_runtime": 0.132,
      "eval_samples_per_second": 2097.784,
      "eval_steps_per_second": 15.146,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.9242841601371765,
      "learning_rate": 4.8e-05,
      "loss": 0.6574,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5342960288808665,
      "eval_loss": 0.7128624320030212,
      "eval_runtime": 0.1516,
      "eval_samples_per_second": 1826.716,
      "eval_steps_per_second": 13.189,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.1665536165237427,
      "learning_rate": 4.7e-05,
      "loss": 0.6158,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.48736462093862815,
      "eval_loss": 0.7406468391418457,
      "eval_runtime": 0.1327,
      "eval_samples_per_second": 2087.06,
      "eval_steps_per_second": 15.069,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.212820291519165,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.538,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5234657039711191,
      "eval_loss": 0.7981071472167969,
      "eval_runtime": 0.1335,
      "eval_samples_per_second": 2075.038,
      "eval_steps_per_second": 14.982,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.8345348834991455,
      "learning_rate": 4.5e-05,
      "loss": 0.4361,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5090252707581228,
      "eval_loss": 0.9250602722167969,
      "eval_runtime": 0.1307,
      "eval_samples_per_second": 2119.955,
      "eval_steps_per_second": 15.307,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.2959797382354736,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.3292,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.51985559566787,
      "eval_loss": 1.1220024824142456,
      "eval_runtime": 0.1286,
      "eval_samples_per_second": 2153.221,
      "eval_steps_per_second": 15.547,
      "step": 60
    },
    {
      "epoch": 6.0,
      "step": 60,
      "total_flos": 989531467960320.0,
      "train_loss": 0.5467174212137859,
      "train_runtime": 22.534,
      "train_samples_per_second": 5524.984,
      "train_steps_per_second": 22.189
    }
  ],
  "logging_steps": 1,
  "max_steps": 500,
  "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": 989531467960320.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}