trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.45887690782546997,
  "best_model_checkpoint": "tiny_bert_km_10_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": 7.391770839691162,
      "learning_rate": 4.9e-05,
      "loss": 0.4317,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7970183486238532,
      "eval_loss": 0.45887690782546997,
      "eval_runtime": 0.268,
      "eval_samples_per_second": 3253.559,
      "eval_steps_per_second": 14.925,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 6.9007887840271,
      "learning_rate": 4.8e-05,
      "loss": 0.2421,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7970183486238532,
      "eval_loss": 0.48968738317489624,
      "eval_runtime": 0.2627,
      "eval_samples_per_second": 3319.372,
      "eval_steps_per_second": 15.226,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.422306060791016,
      "learning_rate": 4.7e-05,
      "loss": 0.1877,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8038990825688074,
      "eval_loss": 0.5509575009346008,
      "eval_runtime": 0.2581,
      "eval_samples_per_second": 3378.838,
      "eval_steps_per_second": 15.499,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 6.809893608093262,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1548,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.7855504587155964,
      "eval_loss": 0.6198525428771973,
      "eval_runtime": 0.2588,
      "eval_samples_per_second": 3369.391,
      "eval_steps_per_second": 15.456,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 6.50020170211792,
      "learning_rate": 4.5e-05,
      "loss": 0.1267,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7752293577981652,
      "eval_loss": 0.6627919673919678,
      "eval_runtime": 0.2679,
      "eval_samples_per_second": 3255.076,
      "eval_steps_per_second": 14.932,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 9.388545036315918,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.1068,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7809633027522935,
      "eval_loss": 0.7954366207122803,
      "eval_runtime": 0.2596,
      "eval_samples_per_second": 3358.83,
      "eval_steps_per_second": 15.407,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 1.0596790573590528e+16,
      "train_loss": 0.20829129219055176,
      "train_runtime": 219.2404,
      "train_samples_per_second": 15359.624,
      "train_steps_per_second": 60.208
    }
  ],
  "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
}