trainer_state.json

{
  "best_global_step": 20,
  "best_metric": 0.6897034049034119,
  "best_model_checkpoint": "distilbert_km_100_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": 1.1397181749343872,
      "learning_rate": 4.9e-05,
      "loss": 0.7134,
      "step": 10
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5234657039711191,
      "eval_loss": 0.7086106538772583,
      "eval_runtime": 0.1346,
      "eval_samples_per_second": 2057.24,
      "eval_steps_per_second": 14.854,
      "step": 10
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.190429449081421,
      "learning_rate": 4.8e-05,
      "loss": 0.6735,
      "step": 20
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.5487364620938628,
      "eval_loss": 0.6897034049034119,
      "eval_runtime": 0.1327,
      "eval_samples_per_second": 2087.311,
      "eval_steps_per_second": 15.071,
      "step": 20
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.9537066221237183,
      "learning_rate": 4.7e-05,
      "loss": 0.6293,
      "step": 30
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.5451263537906137,
      "eval_loss": 0.7048947811126709,
      "eval_runtime": 0.1368,
      "eval_samples_per_second": 2024.994,
      "eval_steps_per_second": 14.621,
      "step": 30
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.0154390335083008,
      "learning_rate": 4.600000000000001e-05,
      "loss": 0.579,
      "step": 40
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.5740072202166066,
      "eval_loss": 0.7152756452560425,
      "eval_runtime": 0.1551,
      "eval_samples_per_second": 1786.174,
      "eval_steps_per_second": 12.897,
      "step": 40
    },
    {
      "epoch": 5.0,
      "grad_norm": 1.1451412439346313,
      "learning_rate": 4.5e-05,
      "loss": 0.5076,
      "step": 50
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.5776173285198556,
      "eval_loss": 0.7509254217147827,
      "eval_runtime": 0.1345,
      "eval_samples_per_second": 2059.366,
      "eval_steps_per_second": 14.869,
      "step": 50
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.7442041635513306,
      "learning_rate": 4.4000000000000006e-05,
      "loss": 0.408,
      "step": 60
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.5270758122743683,
      "eval_loss": 0.8712850213050842,
      "eval_runtime": 0.136,
      "eval_samples_per_second": 2036.188,
      "eval_steps_per_second": 14.702,
      "step": 60
    },
    {
      "epoch": 7.0,
      "grad_norm": 1.9182441234588623,
      "learning_rate": 4.3e-05,
      "loss": 0.2914,
      "step": 70
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.5415162454873647,
      "eval_loss": 0.9978622198104858,
      "eval_runtime": 0.1375,
      "eval_samples_per_second": 2015.153,
      "eval_steps_per_second": 14.55,
      "step": 70
    },
    {
      "epoch": 7.0,
      "step": 70,
      "total_flos": 1154453379287040.0,
      "train_loss": 0.5431749616350446,
      "train_runtime": 26.4729,
      "train_samples_per_second": 4702.926,
      "train_steps_per_second": 18.887
    }
  ],
  "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
}