trainer_state.json

{
  "best_global_step": 23,
  "best_metric": 2.225278854370117,
  "best_model_checkpoint": "tiny_bert_rand_20_v1_stsb/checkpoint-23",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 138,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 10.341554641723633,
      "learning_rate": 4.9e-05,
      "loss": 3.3912,
      "step": 23
    },
    {
      "epoch": 1.0,
      "eval_combined_score": 0.11742589155637043,
      "eval_loss": 2.225278854370117,
      "eval_pearson": 0.12497857622984128,
      "eval_runtime": 0.4625,
      "eval_samples_per_second": 3243.388,
      "eval_spearmanr": 0.10987320688289957,
      "eval_steps_per_second": 12.974,
      "step": 23
    },
    {
      "epoch": 2.0,
      "grad_norm": 6.209749221801758,
      "learning_rate": 4.8e-05,
      "loss": 2.0454,
      "step": 46
    },
    {
      "epoch": 2.0,
      "eval_combined_score": 0.09966272607457684,
      "eval_loss": 2.6571903228759766,
      "eval_pearson": 0.1080683161104216,
      "eval_runtime": 0.4957,
      "eval_samples_per_second": 3026.061,
      "eval_spearmanr": 0.09125713603873208,
      "eval_steps_per_second": 12.104,
      "step": 46
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.432772636413574,
      "learning_rate": 4.7e-05,
      "loss": 1.8227,
      "step": 69
    },
    {
      "epoch": 3.0,
      "eval_combined_score": 0.1716858093451299,
      "eval_loss": 2.375758171081543,
      "eval_pearson": 0.17960902562737494,
      "eval_runtime": 0.4884,
      "eval_samples_per_second": 3071.516,
      "eval_spearmanr": 0.16376259306288485,
      "eval_steps_per_second": 12.286,
      "step": 69
    },
    {
      "epoch": 4.0,
      "grad_norm": 6.799037933349609,
      "learning_rate": 4.600000000000001e-05,
      "loss": 1.6031,
      "step": 92
    },
    {
      "epoch": 4.0,
      "eval_combined_score": 0.23814275891191994,
      "eval_loss": 2.4631683826446533,
      "eval_pearson": 0.24191407857031005,
      "eval_runtime": 0.478,
      "eval_samples_per_second": 3138.068,
      "eval_spearmanr": 0.23437143925352982,
      "eval_steps_per_second": 12.552,
      "step": 92
    },
    {
      "epoch": 5.0,
      "grad_norm": 10.835053443908691,
      "learning_rate": 4.5e-05,
      "loss": 1.2975,
      "step": 115
    },
    {
      "epoch": 5.0,
      "eval_combined_score": 0.27566477480454066,
      "eval_loss": 2.430213212966919,
      "eval_pearson": 0.27735070545682966,
      "eval_runtime": 0.4664,
      "eval_samples_per_second": 3216.377,
      "eval_spearmanr": 0.2739788441522516,
      "eval_steps_per_second": 12.866,
      "step": 115
    },
    {
      "epoch": 6.0,
      "grad_norm": 8.023301124572754,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 1.028,
      "step": 138
    },
    {
      "epoch": 6.0,
      "eval_combined_score": 0.28613560266248894,
      "eval_loss": 2.5196170806884766,
      "eval_pearson": 0.28598421625786247,
      "eval_runtime": 0.4694,
      "eval_samples_per_second": 3195.397,
      "eval_spearmanr": 0.28628698906711547,
      "eval_steps_per_second": 12.782,
      "step": 138
    },
    {
      "epoch": 6.0,
      "step": 138,
      "total_flos": 904528923227136.0,
      "train_loss": 1.8646601248478545,
      "train_runtime": 26.4109,
      "train_samples_per_second": 10883.755,
      "train_steps_per_second": 43.543
    }
  ],
  "logging_steps": 1,
  "max_steps": 1150,
  "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": 904528923227136.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}