trainer_state.json

{
  "best_metric": 0.8901333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0495/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": Infinity,
      "learning_rate": 2.85232922088503e-05,
      "loss": 2.149,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7786666666666666,
      "eval_loss": 0.9604110717773438,
      "eval_runtime": 10.2108,
      "eval_samples_per_second": 367.258,
      "eval_steps_per_second": 5.778,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 15.112010955810547,
      "learning_rate": 2.4368144166557088e-05,
      "loss": 0.6362,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.852,
      "eval_loss": 0.5480005741119385,
      "eval_runtime": 10.225,
      "eval_samples_per_second": 366.747,
      "eval_steps_per_second": 5.77,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 2.4700169563293457,
      "learning_rate": 1.8357520279328874e-05,
      "loss": 0.3192,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8674666666666667,
      "eval_loss": 0.45717164874076843,
      "eval_runtime": 10.523,
      "eval_samples_per_second": 356.363,
      "eval_steps_per_second": 5.607,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.4954148530960083,
      "learning_rate": 1.1681898323495617e-05,
      "loss": 0.1693,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8762666666666666,
      "eval_loss": 0.4176720976829529,
      "eval_runtime": 10.4958,
      "eval_samples_per_second": 357.286,
      "eval_steps_per_second": 5.621,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 2.6349544525146484,
      "learning_rate": 5.6634670985405755e-06,
      "loss": 0.0862,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8882666666666666,
      "eval_loss": 0.392412006855011,
      "eval_runtime": 10.2303,
      "eval_samples_per_second": 366.556,
      "eval_steps_per_second": 5.767,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 1.9008854627609253,
      "learning_rate": 1.4942507197812133e-06,
      "loss": 0.047,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8882666666666666,
      "eval_loss": 0.39207643270492554,
      "eval_runtime": 10.1003,
      "eval_samples_per_second": 371.275,
      "eval_steps_per_second": 5.841,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.5897794961929321,
      "learning_rate": 1.362311416175377e-11,
      "loss": 0.034,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8901333333333333,
      "eval_loss": 0.38807734847068787,
      "eval_runtime": 10.7273,
      "eval_samples_per_second": 349.575,
      "eval_steps_per_second": 5.5,
      "step": 2331
    }
  ],
  "logging_steps": 500,
  "max_steps": 2331,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 7,
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}