trainer_state.json

{
  "best_metric": 0.8474666666666667,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0997/checkpoints/checkpoint-2664",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 2664,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 71.7905502319336,
      "learning_rate": 4.375e-05,
      "loss": 3.7203,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.49333333333333335,
      "eval_loss": 3.224731683731079,
      "eval_runtime": 8.4538,
      "eval_samples_per_second": 443.59,
      "eval_steps_per_second": 6.979,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 198.85562133789062,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 2.0959,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7098666666666666,
      "eval_loss": 1.2014648914337158,
      "eval_runtime": 8.4316,
      "eval_samples_per_second": 444.755,
      "eval_steps_per_second": 6.997,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 199.808837890625,
      "learning_rate": 3.125e-05,
      "loss": 0.9894,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7765333333333333,
      "eval_loss": 0.8225293755531311,
      "eval_runtime": 8.4281,
      "eval_samples_per_second": 444.941,
      "eval_steps_per_second": 7.0,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 340.3732604980469,
      "learning_rate": 2.5e-05,
      "loss": 0.7201,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8098666666666666,
      "eval_loss": 0.6562439203262329,
      "eval_runtime": 8.5276,
      "eval_samples_per_second": 439.75,
      "eval_steps_per_second": 6.919,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 278.65850830078125,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.595,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8301333333333333,
      "eval_loss": 0.572363555431366,
      "eval_runtime": 8.5451,
      "eval_samples_per_second": 438.847,
      "eval_steps_per_second": 6.905,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 151.15374755859375,
      "learning_rate": 1.25e-05,
      "loss": 0.5288,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8445333333333334,
      "eval_loss": 0.5449414253234863,
      "eval_runtime": 9.4158,
      "eval_samples_per_second": 398.266,
      "eval_steps_per_second": 6.266,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 156.1761016845703,
      "learning_rate": 6.25e-06,
      "loss": 0.4913,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8378666666666666,
      "eval_loss": 0.5544743537902832,
      "eval_runtime": 8.2975,
      "eval_samples_per_second": 451.946,
      "eval_steps_per_second": 7.111,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 262.4924011230469,
      "learning_rate": 0.0,
      "loss": 0.4608,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8474666666666667,
      "eval_loss": 0.5187395811080933,
      "eval_runtime": 8.3641,
      "eval_samples_per_second": 448.344,
      "eval_steps_per_second": 7.054,
      "step": 2664
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 6.5411433916416e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}