trainer_state.json

{
  "best_metric": 0.8613333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0378/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": 9e-05,
      "loss": 1.7043,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.772,
      "eval_loss": 0.7826473712921143,
      "eval_runtime": 7.0335,
      "eval_samples_per_second": 533.164,
      "eval_steps_per_second": 8.388,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 1.699253797531128,
      "learning_rate": 9e-05,
      "loss": 0.5812,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.824,
      "eval_loss": 0.5846204161643982,
      "eval_runtime": 6.9014,
      "eval_samples_per_second": 543.366,
      "eval_steps_per_second": 8.549,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": Infinity,
      "learning_rate": 9e-05,
      "loss": 0.3922,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8256,
      "eval_loss": 0.6023015379905701,
      "eval_runtime": 6.6722,
      "eval_samples_per_second": 562.035,
      "eval_steps_per_second": 8.843,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 8.670588493347168,
      "learning_rate": 9e-05,
      "loss": 0.2804,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.8349333333333333,
      "eval_loss": 0.5429110527038574,
      "eval_runtime": 6.7388,
      "eval_samples_per_second": 556.48,
      "eval_steps_per_second": 8.755,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.5022792220115662,
      "learning_rate": 9e-05,
      "loss": 0.2125,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8528,
      "eval_loss": 0.5052595138549805,
      "eval_runtime": 6.5034,
      "eval_samples_per_second": 576.622,
      "eval_steps_per_second": 9.072,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 14.98933219909668,
      "learning_rate": 9e-05,
      "loss": 0.1651,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8613333333333333,
      "eval_loss": 0.48481881618499756,
      "eval_runtime": 7.0237,
      "eval_samples_per_second": 533.906,
      "eval_steps_per_second": 8.4,
      "step": 1998
    }
  ],
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 9.88447923735552e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}