trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.42540276050567627,
  "best_model_checkpoint": "bert_base_km_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": 9.303242683410645,
      "learning_rate": 4.9e-05,
      "loss": 0.4107,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8107798165137615,
      "eval_loss": 0.42540276050567627,
      "eval_runtime": 0.5679,
      "eval_samples_per_second": 1535.37,
      "eval_steps_per_second": 7.043,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.910004615783691,
      "learning_rate": 4.8e-05,
      "loss": 0.2276,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7958715596330275,
      "eval_loss": 0.5113568305969238,
      "eval_runtime": 0.5789,
      "eval_samples_per_second": 1506.374,
      "eval_steps_per_second": 6.91,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.878065586090088,
      "learning_rate": 4.7e-05,
      "loss": 0.1649,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7935779816513762,
      "eval_loss": 0.5798671245574951,
      "eval_runtime": 0.5591,
      "eval_samples_per_second": 1559.586,
      "eval_steps_per_second": 7.154,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 5.366719722747803,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1222,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8142201834862385,
      "eval_loss": 0.6137495636940002,
      "eval_runtime": 0.565,
      "eval_samples_per_second": 1543.303,
      "eval_steps_per_second": 7.079,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.174738645553589,
      "learning_rate": 4.5e-05,
      "loss": 0.0908,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8004587155963303,
      "eval_loss": 0.7026001811027527,
      "eval_runtime": 0.5804,
      "eval_samples_per_second": 1502.323,
      "eval_steps_per_second": 6.891,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 8.232381820678711,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.068,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7981651376146789,
      "eval_loss": 0.8688439726829529,
      "eval_runtime": 0.5657,
      "eval_samples_per_second": 1541.445,
      "eval_steps_per_second": 7.071,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 5.316079940232192e+16,
      "train_loss": 0.18067710688619903,
      "train_runtime": 660.8077,
      "train_samples_per_second": 5095.961,
      "train_steps_per_second": 19.976
    }
  ],
  "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
}