trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.45386210083961487,
  "best_model_checkpoint": "tiny_bert_km_5_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": 8.290016174316406,
      "learning_rate": 4.9e-05,
      "loss": 0.4117,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8061926605504587,
      "eval_loss": 0.45386210083961487,
      "eval_runtime": 0.2584,
      "eval_samples_per_second": 3374.869,
      "eval_steps_per_second": 15.481,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 7.698122501373291,
      "learning_rate": 4.8e-05,
      "loss": 0.237,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8061926605504587,
      "eval_loss": 0.5016503930091858,
      "eval_runtime": 0.2588,
      "eval_samples_per_second": 3369.356,
      "eval_steps_per_second": 15.456,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 5.97835111618042,
      "learning_rate": 4.7e-05,
      "loss": 0.1818,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8107798165137615,
      "eval_loss": 0.5189591646194458,
      "eval_runtime": 0.2633,
      "eval_samples_per_second": 3312.376,
      "eval_steps_per_second": 15.194,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 8.256962776184082,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1468,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8027522935779816,
      "eval_loss": 0.5854922533035278,
      "eval_runtime": 0.258,
      "eval_samples_per_second": 3380.193,
      "eval_steps_per_second": 15.505,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 3.842134952545166,
      "learning_rate": 4.5e-05,
      "loss": 0.1172,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.7958715596330275,
      "eval_loss": 0.6913995146751404,
      "eval_runtime": 0.2644,
      "eval_samples_per_second": 3297.846,
      "eval_steps_per_second": 15.128,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 7.8733086585998535,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0983,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.7970183486238532,
      "eval_loss": 0.6562204360961914,
      "eval_runtime": 0.2583,
      "eval_samples_per_second": 3376.299,
      "eval_steps_per_second": 15.488,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 1.0596790573590528e+16,
      "train_loss": 0.1987890534930759,
      "train_runtime": 219.6846,
      "train_samples_per_second": 15328.568,
      "train_steps_per_second": 60.086
    }
  ],
  "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
}