trainer_state.json

{
  "best_metric": 0.3888,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0507/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": 18.971210479736328,
      "learning_rate": 0.0004294294294294294,
      "loss": 3.5505,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.1648,
      "eval_loss": 3.253709316253662,
      "eval_runtime": 7.6689,
      "eval_samples_per_second": 488.99,
      "eval_steps_per_second": 7.693,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 12.817221641540527,
      "learning_rate": 0.000358000858000858,
      "loss": 3.0443,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2048,
      "eval_loss": 3.0173637866973877,
      "eval_runtime": 7.7784,
      "eval_samples_per_second": 482.107,
      "eval_steps_per_second": 7.585,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 18.8770809173584,
      "learning_rate": 0.0002865722865722866,
      "loss": 2.8489,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.2544,
      "eval_loss": 2.820695400238037,
      "eval_runtime": 9.8611,
      "eval_samples_per_second": 380.281,
      "eval_steps_per_second": 5.983,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 18.830350875854492,
      "learning_rate": 0.00021514371514371513,
      "loss": 2.6668,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.2848,
      "eval_loss": 2.6199207305908203,
      "eval_runtime": 7.6895,
      "eval_samples_per_second": 487.679,
      "eval_steps_per_second": 7.673,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 22.255603790283203,
      "learning_rate": 0.00014371514371514372,
      "loss": 2.4907,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.3336,
      "eval_loss": 2.456575632095337,
      "eval_runtime": 7.8039,
      "eval_samples_per_second": 480.53,
      "eval_steps_per_second": 7.56,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 30.41533851623535,
      "learning_rate": 7.228657228657229e-05,
      "loss": 2.2925,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.37546666666666667,
      "eval_loss": 2.3144500255584717,
      "eval_runtime": 7.7146,
      "eval_samples_per_second": 486.094,
      "eval_steps_per_second": 7.648,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 15.388129234313965,
      "learning_rate": 8.580008580008581e-07,
      "loss": 2.0903,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.3888,
      "eval_loss": 2.2629659175872803,
      "eval_runtime": 8.747,
      "eval_samples_per_second": 428.721,
      "eval_steps_per_second": 6.745,
      "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
}