checkpoint-80/trainer_state.json

{
  "best_metric": 1.2677136659622192,
  "best_model_checkpoint": "/kaggle/working/checkpoint-80",
  "epoch": 8.88888888888889,
  "eval_steps": 10,
  "global_step": 80,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.1111111111111112,
      "grad_norm": 0.02217627689242363,
      "learning_rate": 0.00017777777777777779,
      "loss": 2.0442,
      "step": 10
    },
    {
      "epoch": 1.1111111111111112,
      "eval_loss": 1.737181544303894,
      "eval_runtime": 35.1318,
      "eval_samples_per_second": 1.025,
      "eval_steps_per_second": 0.142,
      "step": 10
    },
    {
      "epoch": 2.2222222222222223,
      "grad_norm": 0.0346713550388813,
      "learning_rate": 0.00015555555555555556,
      "loss": 1.6131,
      "step": 20
    },
    {
      "epoch": 2.2222222222222223,
      "eval_loss": 1.5489343404769897,
      "eval_runtime": 34.8402,
      "eval_samples_per_second": 1.033,
      "eval_steps_per_second": 0.144,
      "step": 20
    },
    {
      "epoch": 3.3333333333333335,
      "grad_norm": 0.02501535415649414,
      "learning_rate": 0.00013333333333333334,
      "loss": 1.4152,
      "step": 30
    },
    {
      "epoch": 3.3333333333333335,
      "eval_loss": 1.4295110702514648,
      "eval_runtime": 34.8537,
      "eval_samples_per_second": 1.033,
      "eval_steps_per_second": 0.143,
      "step": 30
    },
    {
      "epoch": 4.444444444444445,
      "grad_norm": 0.02104916237294674,
      "learning_rate": 0.00011111111111111112,
      "loss": 1.3068,
      "step": 40
    },
    {
      "epoch": 4.444444444444445,
      "eval_loss": 1.3598744869232178,
      "eval_runtime": 35.0281,
      "eval_samples_per_second": 1.028,
      "eval_steps_per_second": 0.143,
      "step": 40
    },
    {
      "epoch": 5.555555555555555,
      "grad_norm": 0.022395364940166473,
      "learning_rate": 8.888888888888889e-05,
      "loss": 1.2049,
      "step": 50
    },
    {
      "epoch": 5.555555555555555,
      "eval_loss": 1.3168741464614868,
      "eval_runtime": 34.692,
      "eval_samples_per_second": 1.038,
      "eval_steps_per_second": 0.144,
      "step": 50
    },
    {
      "epoch": 6.666666666666667,
      "grad_norm": 0.02603345364332199,
      "learning_rate": 6.666666666666667e-05,
      "loss": 1.1086,
      "step": 60
    },
    {
      "epoch": 6.666666666666667,
      "eval_loss": 1.2939578294754028,
      "eval_runtime": 34.6444,
      "eval_samples_per_second": 1.039,
      "eval_steps_per_second": 0.144,
      "step": 60
    },
    {
      "epoch": 7.777777777777778,
      "grad_norm": 0.02798735536634922,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 1.0716,
      "step": 70
    },
    {
      "epoch": 7.777777777777778,
      "eval_loss": 1.2787070274353027,
      "eval_runtime": 34.7046,
      "eval_samples_per_second": 1.037,
      "eval_steps_per_second": 0.144,
      "step": 70
    },
    {
      "epoch": 8.88888888888889,
      "grad_norm": 0.028509726747870445,
      "learning_rate": 2.2222222222222223e-05,
      "loss": 1.0051,
      "step": 80
    },
    {
      "epoch": 8.88888888888889,
      "eval_loss": 1.2677136659622192,
      "eval_runtime": 34.6744,
      "eval_samples_per_second": 1.038,
      "eval_steps_per_second": 0.144,
      "step": 80
    }
  ],
  "logging_steps": 10,
  "max_steps": 90,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 10,
  "stateful_callbacks": {
    "EarlyStoppingCallback": {
      "args": {
        "early_stopping_patience": 3,
        "early_stopping_threshold": 0.0
      },
      "attributes": {
        "early_stopping_patience_counter": 0
      }
    },
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": false
      },
      "attributes": {}
    }
  },
  "total_flos": 2.137153352269824e+16,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}