trainer_state.json

{
  "best_metric": 0.8669333333333333,
  "best_model_checkpoint": "./resnet_finetuned_models_dataset/CIFAR100/50_from_100/microsoft_resnet-101/model_idx_0591/checkpoints/checkpoint-2331",
  "epoch": 7.0,
  "eval_steps": 500,
  "global_step": 2331,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 63.398006439208984,
      "learning_rate": 4.8096988312782174e-05,
      "loss": 3.6444,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.5576,
      "eval_loss": 2.9893767833709717,
      "eval_runtime": 14.5122,
      "eval_samples_per_second": 258.404,
      "eval_steps_per_second": 4.066,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 109.2890853881836,
      "learning_rate": 4.267766952966369e-05,
      "loss": 1.6146,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7848,
      "eval_loss": 0.8186774849891663,
      "eval_runtime": 14.5398,
      "eval_samples_per_second": 257.912,
      "eval_steps_per_second": 4.058,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 203.0039520263672,
      "learning_rate": 3.456708580912725e-05,
      "loss": 0.7183,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8352,
      "eval_loss": 0.5765037536621094,
      "eval_runtime": 14.5554,
      "eval_samples_per_second": 257.637,
      "eval_steps_per_second": 4.053,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 88.63082885742188,
      "learning_rate": 2.5e-05,
      "loss": 0.5256,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8514666666666667,
      "eval_loss": 0.5064903497695923,
      "eval_runtime": 14.611,
      "eval_samples_per_second": 256.656,
      "eval_steps_per_second": 4.038,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 273.61041259765625,
      "learning_rate": 1.5432914190872757e-05,
      "loss": 0.4519,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8613333333333333,
      "eval_loss": 0.46706700325012207,
      "eval_runtime": 14.5785,
      "eval_samples_per_second": 257.229,
      "eval_steps_per_second": 4.047,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 140.36773681640625,
      "learning_rate": 7.3223304703363135e-06,
      "loss": 0.406,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8573333333333333,
      "eval_loss": 0.46818822622299194,
      "eval_runtime": 14.5549,
      "eval_samples_per_second": 257.645,
      "eval_steps_per_second": 4.054,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 90.49237823486328,
      "learning_rate": 1.9030116872178316e-06,
      "loss": 0.3739,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8669333333333333,
      "eval_loss": 0.4470809996128082,
      "eval_runtime": 14.5256,
      "eval_samples_per_second": 258.165,
      "eval_steps_per_second": 4.062,
      "step": 2331
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 5.7235004676864e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}