trainer_state.json

{
  "best_metric": 0.2832,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0089/checkpoints/checkpoint-1665",
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 1665,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 18.939836502075195,
      "learning_rate": 0.0005,
      "loss": 3.4812,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.1704,
      "eval_loss": 3.1859426498413086,
      "eval_runtime": 10.4233,
      "eval_samples_per_second": 359.77,
      "eval_steps_per_second": 5.66,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 16.020660400390625,
      "learning_rate": 0.0005,
      "loss": 3.0274,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.19413333333333332,
      "eval_loss": 3.043713092803955,
      "eval_runtime": 10.282,
      "eval_samples_per_second": 364.716,
      "eval_steps_per_second": 5.738,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 19.708431243896484,
      "learning_rate": 0.0005,
      "loss": 2.8859,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.2456,
      "eval_loss": 2.853959798812866,
      "eval_runtime": 10.2754,
      "eval_samples_per_second": 364.948,
      "eval_steps_per_second": 5.742,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 16.86046600341797,
      "learning_rate": 0.0005,
      "loss": 2.767,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.2608,
      "eval_loss": 2.757610321044922,
      "eval_runtime": 13.5924,
      "eval_samples_per_second": 275.889,
      "eval_steps_per_second": 4.341,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 17.038427352905273,
      "learning_rate": 0.0005,
      "loss": 2.6637,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.2832,
      "eval_loss": 2.6718244552612305,
      "eval_runtime": 10.2595,
      "eval_samples_per_second": 365.514,
      "eval_steps_per_second": 5.751,
      "step": 1665
    }
  ],
  "logging_steps": 500,
  "max_steps": 1665,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 8.2370660311296e+18,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}