trainer_state.json

{
  "best_metric": 0.4712,
  "best_model_checkpoint": "./vit_finetuned_models_dataset/CIFAR100/50_from_100/facebook_dino-vitb16/model_idx_0321/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": 0.0002911145304096916,
      "loss": 3.3757,
      "step": 333
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.19626666666666667,
      "eval_loss": 3.035726547241211,
      "eval_runtime": 5.8496,
      "eval_samples_per_second": 641.069,
      "eval_steps_per_second": 10.086,
      "step": 333
    },
    {
      "epoch": 2.0,
      "grad_norm": 22.072351455688477,
      "learning_rate": 0.00026520930752569493,
      "loss": 2.7756,
      "step": 666
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.2946666666666667,
      "eval_loss": 2.6711208820343018,
      "eval_runtime": 5.9072,
      "eval_samples_per_second": 634.816,
      "eval_steps_per_second": 9.988,
      "step": 666
    },
    {
      "epoch": 3.0,
      "grad_norm": 18.774932861328125,
      "learning_rate": 0.00022540814184580826,
      "loss": 2.4882,
      "step": 999
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.332,
      "eval_loss": 2.471115827560425,
      "eval_runtime": 5.9341,
      "eval_samples_per_second": 631.937,
      "eval_steps_per_second": 9.942,
      "step": 999
    },
    {
      "epoch": 4.0,
      "grad_norm": 39.158573150634766,
      "learning_rate": 0.00017651164135367125,
      "loss": 2.274,
      "step": 1332
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.3738666666666667,
      "eval_loss": 2.3249166011810303,
      "eval_runtime": 5.8969,
      "eval_samples_per_second": 635.923,
      "eval_steps_per_second": 10.005,
      "step": 1332
    },
    {
      "epoch": 5.0,
      "grad_norm": 31.92185401916504,
      "learning_rate": 0.00012441744564412654,
      "loss": 2.0665,
      "step": 1665
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.4096,
      "eval_loss": 2.1682546138763428,
      "eval_runtime": 7.0747,
      "eval_samples_per_second": 530.056,
      "eval_steps_per_second": 8.34,
      "step": 1665
    },
    {
      "epoch": 6.0,
      "grad_norm": 25.809804916381836,
      "learning_rate": 7.540888354819138e-05,
      "loss": 1.8324,
      "step": 1998
    },
    {
      "epoch": 6.0,
      "eval_accuracy": 0.4322666666666667,
      "eval_loss": 2.0662786960601807,
      "eval_runtime": 7.0505,
      "eval_samples_per_second": 531.875,
      "eval_steps_per_second": 8.368,
      "step": 1998
    },
    {
      "epoch": 7.0,
      "grad_norm": 35.65645217895508,
      "learning_rate": 3.5397110943979596e-05,
      "loss": 1.584,
      "step": 2331
    },
    {
      "epoch": 7.0,
      "eval_accuracy": 0.4594666666666667,
      "eval_loss": 2.0200963020324707,
      "eval_runtime": 7.2539,
      "eval_samples_per_second": 516.965,
      "eval_steps_per_second": 8.134,
      "step": 2331
    },
    {
      "epoch": 8.0,
      "grad_norm": Infinity,
      "learning_rate": 9.262458019425962e-06,
      "loss": 1.3706,
      "step": 2664
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.4712,
      "eval_loss": 2.0176968574523926,
      "eval_runtime": 5.9656,
      "eval_samples_per_second": 628.603,
      "eval_steps_per_second": 9.89,
      "step": 2664
    }
  ],
  "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": 1.317930564980736e+19,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}