trainer_state.json

{
  "best_global_step": 264,
  "best_metric": 0.43114957213401794,
  "best_model_checkpoint": "bert_base_km_100_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": 8.772541999816895,
      "learning_rate": 4.9e-05,
      "loss": 0.4229,
      "step": 264
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.8061926605504587,
      "eval_loss": 0.43114957213401794,
      "eval_runtime": 0.5927,
      "eval_samples_per_second": 1471.301,
      "eval_steps_per_second": 6.749,
      "step": 264
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.649755477905273,
      "learning_rate": 4.8e-05,
      "loss": 0.2261,
      "step": 528
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7924311926605505,
      "eval_loss": 0.5237594246864319,
      "eval_runtime": 0.5723,
      "eval_samples_per_second": 1523.657,
      "eval_steps_per_second": 6.989,
      "step": 528
    },
    {
      "epoch": 3.0,
      "grad_norm": 4.963180065155029,
      "learning_rate": 4.7e-05,
      "loss": 0.1537,
      "step": 792
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7878440366972477,
      "eval_loss": 0.5836676955223083,
      "eval_runtime": 0.562,
      "eval_samples_per_second": 1551.472,
      "eval_steps_per_second": 7.117,
      "step": 792
    },
    {
      "epoch": 4.0,
      "grad_norm": 9.778557777404785,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.1123,
      "step": 1056
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8061926605504587,
      "eval_loss": 0.554785966873169,
      "eval_runtime": 0.5622,
      "eval_samples_per_second": 1550.924,
      "eval_steps_per_second": 7.114,
      "step": 1056
    },
    {
      "epoch": 5.0,
      "grad_norm": 10.721723556518555,
      "learning_rate": 4.5e-05,
      "loss": 0.0796,
      "step": 1320
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8061926605504587,
      "eval_loss": 0.6690517067909241,
      "eval_runtime": 0.5798,
      "eval_samples_per_second": 1503.909,
      "eval_steps_per_second": 6.899,
      "step": 1320
    },
    {
      "epoch": 6.0,
      "grad_norm": 4.865378379821777,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.0588,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.786697247706422,
      "eval_loss": 0.9136205315589905,
      "eval_runtime": 0.5608,
      "eval_samples_per_second": 1554.958,
      "eval_steps_per_second": 7.133,
      "step": 1584
    },
    {
      "epoch": 6.0,
      "step": 1584,
      "total_flos": 5.316079940232192e+16,
      "train_loss": 0.17555388898560495,
      "train_runtime": 662.4893,
      "train_samples_per_second": 5083.026,
      "train_steps_per_second": 19.925
    }
  ],
  "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": 5.316079940232192e+16,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}