trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.44193360209465027,
  "best_model_checkpoint": "tiny_bert_km_5_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": 7.1062188148498535,
      "learning_rate": 4.9e-05,
      "loss": 0.4137,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8130733944954128,
      "eval_loss": 0.44193360209465027,
      "eval_runtime": 0.2752,
      "eval_samples_per_second": 3168.51,
      "eval_steps_per_second": 14.534,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.389181137084961,
      "learning_rate": 4.8e-05,
      "loss": 0.2547,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8107798165137615,
      "eval_loss": 0.47274479269981384,
      "eval_runtime": 0.2612,
      "eval_samples_per_second": 3338.177,
      "eval_steps_per_second": 15.313,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.381705284118652,
      "learning_rate": 4.7e-05,
      "loss": 0.1972,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8107798165137615,
      "eval_loss": 0.4966350197792053,
      "eval_runtime": 0.2609,
      "eval_samples_per_second": 3342.151,
      "eval_steps_per_second": 15.331,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 5.678281307220459,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1623,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8176605504587156,
      "eval_loss": 0.46778005361557007,
      "eval_runtime": 0.2752,
      "eval_samples_per_second": 3168.123,
      "eval_steps_per_second": 14.533,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 6.3953351974487305,
      "learning_rate": 4.5e-05,
      "loss": 0.1359,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8096330275229358,
      "eval_loss": 0.5381088256835938,
      "eval_runtime": 0.2676,
      "eval_samples_per_second": 3259.102,
      "eval_steps_per_second": 14.95,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 3.137275218963623,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.1171,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8142201834862385,
      "eval_loss": 0.6161597967147827,
      "eval_runtime": 0.2605,
      "eval_samples_per_second": 3347.238,
      "eval_steps_per_second": 15.354,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 1.0596790573590528e+16,
      "train_loss": 0.21348130221318717,
      "train_runtime": 215.8494,
      "train_samples_per_second": 15600.923,
      "train_steps_per_second": 61.154
    }
  ],
  "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
}