trainer_state.json

{
  "best_metric": 0.9583333333333334,
  "best_model_checkpoint": "portrait_cosu_exp5/checkpoint-40",
  "epoch": 3.902439024390244,
  "eval_steps": 500,
  "global_step": 40,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.975609756097561,
      "grad_norm": 13.03514289855957,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.3966,
      "step": 10
    },
    {
      "epoch": 0.975609756097561,
      "eval_accuracy": 0.8611111111111112,
      "eval_confusion_matrix": [
        [
          50,
          1
        ],
        [
          9,
          12
        ]
      ],
      "eval_f1": 0.8498217468805704,
      "eval_loss": 0.30243051052093506,
      "eval_precision": 0.8695132551064755,
      "eval_recall": 0.8611111111111112,
      "eval_runtime": 5.8713,
      "eval_samples_per_second": 12.263,
      "eval_steps_per_second": 0.852,
      "step": 10
    },
    {
      "epoch": 1.951219512195122,
      "grad_norm": 18.37648582458496,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.3486,
      "step": 20
    },
    {
      "epoch": 1.951219512195122,
      "eval_accuracy": 0.8333333333333334,
      "eval_confusion_matrix": [
        [
          50,
          1
        ],
        [
          11,
          10
        ]
      ],
      "eval_f1": 0.8147321428571428,
      "eval_loss": 0.3285925090312958,
      "eval_precision": 0.84575260804769,
      "eval_recall": 0.8333333333333334,
      "eval_runtime": 5.962,
      "eval_samples_per_second": 12.077,
      "eval_steps_per_second": 0.839,
      "step": 20
    },
    {
      "epoch": 2.926829268292683,
      "grad_norm": 4.450592517852783,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.2347,
      "step": 30
    },
    {
      "epoch": 2.926829268292683,
      "eval_accuracy": 0.9444444444444444,
      "eval_confusion_matrix": [
        [
          50,
          1
        ],
        [
          3,
          18
        ]
      ],
      "eval_f1": 0.9435897435897436,
      "eval_loss": 0.17685942351818085,
      "eval_precision": 0.9445547831843761,
      "eval_recall": 0.9444444444444444,
      "eval_runtime": 5.6367,
      "eval_samples_per_second": 12.773,
      "eval_steps_per_second": 0.887,
      "step": 30
    },
    {
      "epoch": 3.902439024390244,
      "grad_norm": 6.283059597015381,
      "learning_rate": 0.0,
      "loss": 0.2039,
      "step": 40
    },
    {
      "epoch": 3.902439024390244,
      "eval_accuracy": 0.9583333333333334,
      "eval_confusion_matrix": [
        [
          50,
          1
        ],
        [
          2,
          19
        ]
      ],
      "eval_f1": 0.9580274686242009,
      "eval_loss": 0.16026876866817474,
      "eval_precision": 0.9581730769230768,
      "eval_recall": 0.9583333333333334,
      "eval_runtime": 5.4865,
      "eval_samples_per_second": 13.123,
      "eval_steps_per_second": 0.911,
      "step": 40
    },
    {
      "epoch": 3.902439024390244,
      "step": 40,
      "total_flos": 5.787149236001833e+17,
      "train_loss": 0.2959394633769989,
      "train_runtime": 442.5785,
      "train_samples_per_second": 5.857,
      "train_steps_per_second": 0.09
    }
  ],
  "logging_steps": 10,
  "max_steps": 40,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 5.787149236001833e+17,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}