trainer_state.json

{
  "best_metric": 0.8456,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0361/checkpoints/checkpoint-2997",
  "epoch": 9.0,
  "eval_steps": 500,
  "global_step": 2997,
  "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": 9e-05,
      "loss": 1.9438,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7402666666666666,
      "eval_loss": 0.9024656414985657,
      "eval_runtime": 7.7395,
      "eval_samples_per_second": 484.53,
      "eval_steps_per_second": 7.623,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": Infinity,
      "learning_rate": 9e-05,
      "loss": 0.6536,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8064,
      "eval_loss": 0.6462198495864868,
      "eval_runtime": 6.9567,
      "eval_samples_per_second": 539.052,
      "eval_steps_per_second": 8.481,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 37.62237548828125,
      "learning_rate": 9e-05,
      "loss": 0.3901,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8248,
      "eval_loss": 0.5749683380126953,
      "eval_runtime": 7.3095,
      "eval_samples_per_second": 513.031,
      "eval_steps_per_second": 8.072,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.565747618675232,
      "learning_rate": 9e-05,
      "loss": 0.2559,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8186666666666667,
      "eval_loss": 0.611893355846405,
      "eval_runtime": 6.8787,
      "eval_samples_per_second": 545.163,
      "eval_steps_per_second": 8.577,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.7178195118904114,
      "learning_rate": 9e-05,
      "loss": 0.1771,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8229333333333333,
      "eval_loss": 0.6176028847694397,
      "eval_runtime": 6.9655,
      "eval_samples_per_second": 538.369,
      "eval_steps_per_second": 8.47,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": Infinity,
      "learning_rate": 9e-05,
      "loss": 0.1305,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8392,
      "eval_loss": 0.5892927050590515,
      "eval_runtime": 6.938,
      "eval_samples_per_second": 540.505,
      "eval_steps_per_second": 8.504,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.08140281587839127,
      "learning_rate": 9e-05,
      "loss": 0.1085,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.8418666666666667,
      "eval_loss": 0.5986810922622681,
      "eval_runtime": 7.0054,
      "eval_samples_per_second": 535.3,
      "eval_steps_per_second": 8.422,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.05093102529644966,
      "learning_rate": 9e-05,
      "loss": 0.0884,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.8266666666666667,
      "eval_loss": 0.6526540517807007,
      "eval_runtime": 6.9652,
      "eval_samples_per_second": 538.392,
      "eval_steps_per_second": 8.471,
      "step": 2664
    },
    {
      "epoch": 9.0,
      "grad_norm": 0.0183145459741354,
      "learning_rate": 9e-05,
      "loss": 0.0816,
      "step": 2997
    },
    {
      "epoch": 9.0,
      "eval_accuracy": 0.8456,
      "eval_loss": 0.6241752505302429,
      "eval_runtime": 7.5129,
      "eval_samples_per_second": 499.139,
      "eval_steps_per_second": 7.853,
      "step": 2997
    }
  ],
  "logging_steps": 500,
  "max_steps": 2997,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9,
  "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": 1.482671885603328e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}