trainer_state.json

{
  "best_global_step": 20,
  "best_metric": 0.7060370445251465,
  "best_model_checkpoint": "distilbert_km_5_v2_rte/checkpoint-20",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 70,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 1.596053957939148,
      "learning_rate": 4.9e-05,
      "loss": 0.7078,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5054151624548736,
      "eval_loss": 0.7139623761177063,
      "eval_runtime": 0.1253,
      "eval_samples_per_second": 2210.241,
      "eval_steps_per_second": 15.958,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.9824621677398682,
      "learning_rate": 4.8e-05,
      "loss": 0.6476,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5234657039711191,
      "eval_loss": 0.7060370445251465,
      "eval_runtime": 0.1345,
      "eval_samples_per_second": 2059.388,
      "eval_steps_per_second": 14.869,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.229831337928772,
      "learning_rate": 4.7e-05,
      "loss": 0.5879,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.4981949458483754,
      "eval_loss": 0.7476572394371033,
      "eval_runtime": 0.1428,
      "eval_samples_per_second": 1940.089,
      "eval_steps_per_second": 14.008,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.4849236011505127,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.4946,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.4981949458483754,
      "eval_loss": 0.8227261304855347,
      "eval_runtime": 0.1284,
      "eval_samples_per_second": 2156.635,
      "eval_steps_per_second": 15.571,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.8758838176727295,
      "learning_rate": 4.5e-05,
      "loss": 0.3673,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5090252707581228,
      "eval_loss": 1.0282340049743652,
      "eval_runtime": 0.1337,
      "eval_samples_per_second": 2072.262,
      "eval_steps_per_second": 14.962,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 4.0598039627075195,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.2294,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.49458483754512633,
      "eval_loss": 1.4427438974380493,
      "eval_runtime": 0.1316,
      "eval_samples_per_second": 2104.586,
      "eval_steps_per_second": 15.196,
      "step": 60
    },
    {
      "epoch": 7.0,
      "grad_norm": 2.625215768814087,
      "learning_rate": 4.3e-05,
      "loss": 0.1442,
      "step": 70
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.4693140794223827,
      "eval_loss": 1.594504475593567,
      "eval_runtime": 0.1264,
      "eval_samples_per_second": 2190.844,
      "eval_steps_per_second": 15.818,
      "step": 70
    },
    {
      "epoch": 7.0,
      "step": 70,
      "total_flos": 1154453379287040.0,
      "train_loss": 0.45409974200384956,
      "train_runtime": 26.8997,
      "train_samples_per_second": 4628.306,
      "train_steps_per_second": 18.588
    }
  ],
  "logging_steps": 1,
  "max_steps": 500,
  "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": 1154453379287040.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}