trainer_state.json

{
  "best_metric": 0.9021333333333333,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_vit-mae-base/model_idx_0474/checkpoints/checkpoint-2664",
  "epoch": 8.0,
  "eval_steps": 500,
  "global_step": 2664,
  "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": 6.736728394143519e-05,
      "loss": 1.7455,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7728,
      "eval_loss": 0.7963178157806396,
      "eval_runtime": 7.2075,
      "eval_samples_per_second": 520.291,
      "eval_steps_per_second": 8.186,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 48.0760383605957,
      "learning_rate": 5.9807039661419336e-05,
      "loss": 0.5345,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.8434666666666667,
      "eval_loss": 0.5159484148025513,
      "eval_runtime": 7.3368,
      "eval_samples_per_second": 511.121,
      "eval_steps_per_second": 8.042,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 29.665470123291016,
      "learning_rate": 4.8470148469326895e-05,
      "loss": 0.3003,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.8704,
      "eval_loss": 0.43053963780403137,
      "eval_runtime": 7.4205,
      "eval_samples_per_second": 505.36,
      "eval_steps_per_second": 7.951,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 1.162221074104309,
      "learning_rate": 3.508254927997938e-05,
      "loss": 0.1667,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.88,
      "eval_loss": 0.4175266921520233,
      "eval_runtime": 7.3744,
      "eval_samples_per_second": 508.515,
      "eval_steps_per_second": 8.001,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.146314337849617,
      "learning_rate": 2.1682382711066092e-05,
      "loss": 0.0935,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.8930666666666667,
      "eval_loss": 0.37221816182136536,
      "eval_runtime": 8.031,
      "eval_samples_per_second": 466.94,
      "eval_steps_per_second": 7.347,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.084595687687397,
      "learning_rate": 1.0309702649891636e-05,
      "loss": 0.0525,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.8941333333333333,
      "eval_loss": 0.3724525272846222,
      "eval_runtime": 7.6122,
      "eval_samples_per_second": 492.633,
      "eval_steps_per_second": 7.751,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 0.07316591590642929,
      "learning_rate": 2.6958965421683553e-06,
      "loss": 0.0272,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9010666666666667,
      "eval_loss": 0.369891881942749,
      "eval_runtime": 7.5014,
      "eval_samples_per_second": 499.909,
      "eval_steps_per_second": 7.865,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": 0.064798504114151,
      "learning_rate": 9.734847288167535e-11,
      "loss": 0.0185,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9021333333333333,
      "eval_loss": 0.3596207797527313,
      "eval_runtime": 7.6782,
      "eval_samples_per_second": 488.396,
      "eval_steps_per_second": 7.684,
      "step": 2664
    }
  ],
  "logging_steps": 500,
  "max_steps": 2664,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 8,
  "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": 1.317930564980736e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}