trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.4364543855190277,
  "best_model_checkpoint": "tiny_bert_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": 7.109645843505859,
      "learning_rate": 4.9e-05,
      "loss": 0.4162,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8038990825688074,
      "eval_loss": 0.4364543855190277,
      "eval_runtime": 0.2606,
      "eval_samples_per_second": 3346.429,
      "eval_steps_per_second": 15.351,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.221988677978516,
      "learning_rate": 4.8e-05,
      "loss": 0.2405,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8176605504587156,
      "eval_loss": 0.4643456041812897,
      "eval_runtime": 0.2911,
      "eval_samples_per_second": 2995.127,
      "eval_steps_per_second": 13.739,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.5154852867126465,
      "learning_rate": 4.7e-05,
      "loss": 0.1811,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8153669724770642,
      "eval_loss": 0.48837944865226746,
      "eval_runtime": 0.2995,
      "eval_samples_per_second": 2911.512,
      "eval_steps_per_second": 13.356,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 6.539164066314697,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1447,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8130733944954128,
      "eval_loss": 0.5304074883460999,
      "eval_runtime": 0.2597,
      "eval_samples_per_second": 3358.352,
      "eval_steps_per_second": 15.405,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 6.107559680938721,
      "learning_rate": 4.5e-05,
      "loss": 0.1155,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8096330275229358,
      "eval_loss": 0.6028443574905396,
      "eval_runtime": 0.2607,
      "eval_samples_per_second": 3345.181,
      "eval_steps_per_second": 15.345,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 9.331040382385254,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0925,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8004587155963303,
      "eval_loss": 0.6613821387290955,
      "eval_runtime": 0.2617,
      "eval_samples_per_second": 3332.458,
      "eval_steps_per_second": 15.287,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 1.0596790573590528e+16,
      "train_loss": 0.19841254118717078,
      "train_runtime": 215.9489,
      "train_samples_per_second": 15593.739,
      "train_steps_per_second": 61.126
    }
  ],
  "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
}