trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.4675423502922058,
  "best_model_checkpoint": "tiny_bert_rand_20_v1_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": 6.416019916534424,
      "learning_rate": 4.9e-05,
      "loss": 0.4258,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7901376146788991,
      "eval_loss": 0.4675423502922058,
      "eval_runtime": 0.2656,
      "eval_samples_per_second": 3283.274,
      "eval_steps_per_second": 15.061,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 10.649538040161133,
      "learning_rate": 4.8e-05,
      "loss": 0.2409,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8027522935779816,
      "eval_loss": 0.5330525040626526,
      "eval_runtime": 0.2701,
      "eval_samples_per_second": 3228.324,
      "eval_steps_per_second": 14.809,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.359218597412109,
      "learning_rate": 4.7e-05,
      "loss": 0.1764,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.805045871559633,
      "eval_loss": 0.5286797881126404,
      "eval_runtime": 0.2682,
      "eval_samples_per_second": 3251.476,
      "eval_steps_per_second": 14.915,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 9.089826583862305,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1362,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.805045871559633,
      "eval_loss": 0.6284171342849731,
      "eval_runtime": 0.2672,
      "eval_samples_per_second": 3263.106,
      "eval_steps_per_second": 14.968,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.6520237922668457,
      "learning_rate": 4.5e-05,
      "loss": 0.1034,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7947247706422018,
      "eval_loss": 0.6998069286346436,
      "eval_runtime": 0.2665,
      "eval_samples_per_second": 3272.528,
      "eval_steps_per_second": 15.012,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 6.043944358825684,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0825,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7981651376146789,
      "eval_loss": 0.7145135402679443,
      "eval_runtime": 0.271,
      "eval_samples_per_second": 3217.586,
      "eval_steps_per_second": 14.76,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 1.0596790573590528e+16,
      "train_loss": 0.1941958340731534,
      "train_runtime": 223.2636,
      "train_samples_per_second": 15082.845,
      "train_steps_per_second": 59.123
    }
  ],
  "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": 1.0596790573590528e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}