trainer_state.json

{
  "best_global_step": 10,
  "best_metric": 0.6883178949356079,
  "best_model_checkpoint": "distilbert_rand_100_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": 0.7858245372772217,
      "learning_rate": 4.9e-05,
      "loss": 0.6953,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5270758122743683,
      "eval_loss": 0.6883178949356079,
      "eval_runtime": 0.1611,
      "eval_samples_per_second": 1719.585,
      "eval_steps_per_second": 12.416,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.6864529848098755,
      "learning_rate": 4.8e-05,
      "loss": 0.6821,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5379061371841155,
      "eval_loss": 0.7001636028289795,
      "eval_runtime": 0.1549,
      "eval_samples_per_second": 1788.296,
      "eval_steps_per_second": 12.912,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.6565799713134766,
      "learning_rate": 4.7e-05,
      "loss": 0.6373,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5054151624548736,
      "eval_loss": 0.7812218070030212,
      "eval_runtime": 0.1687,
      "eval_samples_per_second": 1641.88,
      "eval_steps_per_second": 11.855,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 3.382657051086426,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.5267,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5126353790613718,
      "eval_loss": 0.9471554756164551,
      "eval_runtime": 0.1487,
      "eval_samples_per_second": 1862.286,
      "eval_steps_per_second": 13.446,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.9300932884216309,
      "learning_rate": 4.5e-05,
      "loss": 0.3732,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5090252707581228,
      "eval_loss": 1.1737821102142334,
      "eval_runtime": 0.1431,
      "eval_samples_per_second": 1935.661,
      "eval_steps_per_second": 13.976,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 7.640570640563965,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.262,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5018050541516246,
      "eval_loss": 1.6651986837387085,
      "eval_runtime": 0.1301,
      "eval_samples_per_second": 2128.71,
      "eval_steps_per_second": 15.37,
      "step": 60
    },
    {
      "epoch": 6.0,
      "step": 60,
      "total_flos": 989531467960320.0,
      "train_loss": 0.5294327775637309,
      "train_runtime": 23.7523,
      "train_samples_per_second": 5241.608,
      "train_steps_per_second": 21.051
    }
  ],
  "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
}