trainer_state.json

{
  "best_metric": 0.8568,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0042/checkpoints/checkpoint-1998",
  "epoch": 6.0,
  "eval_steps": 500,
  "global_step": 1998,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": Infinity,
      "learning_rate": 0.0001,
      "loss": 1.6425,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7781333333333333,
      "eval_loss": 0.75034499168396,
      "eval_runtime": 13.4616,
      "eval_samples_per_second": 278.571,
      "eval_steps_per_second": 4.383,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.7167515754699707,
      "learning_rate": 0.0001,
      "loss": 0.5486,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8314666666666667,
      "eval_loss": 0.5361385345458984,
      "eval_runtime": 13.0151,
      "eval_samples_per_second": 288.126,
      "eval_steps_per_second": 4.533,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 9.805776596069336,
      "learning_rate": 0.0001,
      "loss": 0.3486,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.84,
      "eval_loss": 0.5276156663894653,
      "eval_runtime": 12.9327,
      "eval_samples_per_second": 289.962,
      "eval_steps_per_second": 4.562,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 4.685305595397949,
      "learning_rate": 0.0001,
      "loss": 0.2391,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8336,
      "eval_loss": 0.5486657619476318,
      "eval_runtime": 13.314,
      "eval_samples_per_second": 281.658,
      "eval_steps_per_second": 4.431,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 20.503311157226562,
      "learning_rate": 0.0001,
      "loss": 0.1723,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8282666666666667,
      "eval_loss": 0.5651600360870361,
      "eval_runtime": 12.6566,
      "eval_samples_per_second": 296.289,
      "eval_steps_per_second": 4.662,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.6760744452476501,
      "learning_rate": 0.0001,
      "loss": 0.1353,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8568,
      "eval_loss": 0.5035337209701538,
      "eval_runtime": 13.308,
      "eval_samples_per_second": 281.786,
      "eval_steps_per_second": 4.433,
      "step": 1998
    }
  ],
  "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": 9.88447923735552e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}