trainer_state.json

{
  "best_metric": 0.9370666666666667,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/google_vit-base-patch16-224/model_idx_0962/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": 0.5772020816802979,
      "learning_rate": 4.849231551964771e-05,
      "loss": 0.8877,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.9136,
      "eval_loss": 0.2971617877483368,
      "eval_runtime": 13.4064,
      "eval_samples_per_second": 279.716,
      "eval_steps_per_second": 4.401,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.331511527299881,
      "learning_rate": 4.415111107797445e-05,
      "loss": 0.1754,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.92,
      "eval_loss": 0.2677749693393707,
      "eval_runtime": 13.4386,
      "eval_samples_per_second": 279.048,
      "eval_steps_per_second": 4.39,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 0.26810869574546814,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.0927,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.9272,
      "eval_loss": 0.2518373429775238,
      "eval_runtime": 14.2952,
      "eval_samples_per_second": 262.325,
      "eval_steps_per_second": 4.127,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 2.0437963008880615,
      "learning_rate": 2.9341204441673266e-05,
      "loss": 0.0501,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.932,
      "eval_loss": 0.2519501745700836,
      "eval_runtime": 13.3391,
      "eval_samples_per_second": 281.129,
      "eval_steps_per_second": 4.423,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 0.016509361565113068,
      "learning_rate": 2.0658795558326743e-05,
      "loss": 0.0238,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.936,
      "eval_loss": 0.24863170087337494,
      "eval_runtime": 13.1226,
      "eval_samples_per_second": 285.766,
      "eval_steps_per_second": 4.496,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 0.011792152188718319,
      "learning_rate": 1.2500000000000006e-05,
      "loss": 0.0105,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.9344,
      "eval_loss": 0.2547927498817444,
      "eval_runtime": 13.1434,
      "eval_samples_per_second": 285.315,
      "eval_steps_per_second": 4.489,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": NaN,
      "learning_rate": 5.865744425999756e-06,
      "loss": 0.006,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.9370666666666667,
      "eval_loss": 0.268070250749588,
      "eval_runtime": 13.2027,
      "eval_samples_per_second": 284.034,
      "eval_steps_per_second": 4.469,
      "step": 2331
    }
  ],
  "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": 1.153189244358144e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}