trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.40931466221809387,
  "best_model_checkpoint": "bert_base_rand_10_v2_sst2/checkpoint-264",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1584,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 8.08377456665039,
      "learning_rate": 4.9e-05,
      "loss": 0.3846,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8130733944954128,
      "eval_loss": 0.40931466221809387,
      "eval_runtime": 0.5732,
      "eval_samples_per_second": 1521.192,
      "eval_steps_per_second": 6.978,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.9197845458984375,
      "learning_rate": 4.8e-05,
      "loss": 0.2226,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7993119266055045,
      "eval_loss": 0.5550717711448669,
      "eval_runtime": 0.5718,
      "eval_samples_per_second": 1525.0,
      "eval_steps_per_second": 6.995,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.202040195465088,
      "learning_rate": 4.7e-05,
      "loss": 0.1669,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8096330275229358,
      "eval_loss": 0.5425371527671814,
      "eval_runtime": 0.5677,
      "eval_samples_per_second": 1536.036,
      "eval_steps_per_second": 7.046,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 6.554544448852539,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.127,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.805045871559633,
      "eval_loss": 0.5878366827964783,
      "eval_runtime": 0.5761,
      "eval_samples_per_second": 1513.509,
      "eval_steps_per_second": 6.943,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 12.496179580688477,
      "learning_rate": 4.5e-05,
      "loss": 0.1007,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7924311926605505,
      "eval_loss": 0.5887973308563232,
      "eval_runtime": 0.565,
      "eval_samples_per_second": 1543.268,
      "eval_steps_per_second": 7.079,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 10.65781021118164,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0807,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7970183486238532,
      "eval_loss": 0.8107673525810242,
      "eval_runtime": 0.5624,
      "eval_samples_per_second": 1550.494,
      "eval_steps_per_second": 7.112,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 5.316079940232192e+16,
      "train_loss": 0.1804122780308579,
      "train_runtime": 667.8738,
      "train_samples_per_second": 5042.046,
      "train_steps_per_second": 19.764
    }
  ],
  "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": 5.316079940232192e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}