checkpoint-160/trainer_state.json

{
  "best_metric": 0.9323529411764706,
  "best_model_checkpoint": "vit_fulldataset/checkpoint-160",
  "epoch": 5.0,
  "global_step": 160,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.31,
      "learning_rate": 2.5e-05,
      "loss": 4.6284,
      "step": 10
    },
    {
      "epoch": 0.62,
      "learning_rate": 5e-05,
      "loss": 4.5942,
      "step": 20
    },
    {
      "epoch": 0.94,
      "learning_rate": 4.709302325581396e-05,
      "loss": 4.5291,
      "step": 30
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.27450980392156865,
      "eval_loss": 4.447311878204346,
      "eval_runtime": 21.4513,
      "eval_samples_per_second": 47.55,
      "eval_steps_per_second": 5.967,
      "step": 32
    },
    {
      "epoch": 1.25,
      "learning_rate": 4.418604651162791e-05,
      "loss": 4.4148,
      "step": 40
    },
    {
      "epoch": 1.56,
      "learning_rate": 4.127906976744187e-05,
      "loss": 4.3267,
      "step": 50
    },
    {
      "epoch": 1.88,
      "learning_rate": 3.837209302325582e-05,
      "loss": 4.2447,
      "step": 60
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7598039215686274,
      "eval_loss": 4.18092679977417,
      "eval_runtime": 21.9573,
      "eval_samples_per_second": 46.454,
      "eval_steps_per_second": 5.829,
      "step": 64
    },
    {
      "epoch": 2.19,
      "learning_rate": 3.5465116279069774e-05,
      "loss": 4.1564,
      "step": 70
    },
    {
      "epoch": 2.5,
      "learning_rate": 3.2558139534883724e-05,
      "loss": 4.0352,
      "step": 80
    },
    {
      "epoch": 2.81,
      "learning_rate": 2.9651162790697678e-05,
      "loss": 3.9991,
      "step": 90
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.884313725490196,
      "eval_loss": 3.975513219833374,
      "eval_runtime": 22.2832,
      "eval_samples_per_second": 45.774,
      "eval_steps_per_second": 5.744,
      "step": 96
    },
    {
      "epoch": 3.12,
      "learning_rate": 2.674418604651163e-05,
      "loss": 3.9439,
      "step": 100
    },
    {
      "epoch": 3.44,
      "learning_rate": 2.3837209302325582e-05,
      "loss": 3.8316,
      "step": 110
    },
    {
      "epoch": 3.75,
      "learning_rate": 2.0930232558139536e-05,
      "loss": 3.8242,
      "step": 120
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9303921568627451,
      "eval_loss": 3.8248324394226074,
      "eval_runtime": 21.9121,
      "eval_samples_per_second": 46.55,
      "eval_steps_per_second": 5.842,
      "step": 128
    },
    {
      "epoch": 4.06,
      "learning_rate": 1.802325581395349e-05,
      "loss": 3.7498,
      "step": 130
    },
    {
      "epoch": 4.38,
      "learning_rate": 1.5116279069767441e-05,
      "loss": 3.702,
      "step": 140
    },
    {
      "epoch": 4.69,
      "learning_rate": 1.2209302325581395e-05,
      "loss": 3.6716,
      "step": 150
    },
    {
      "epoch": 5.0,
      "learning_rate": 9.302325581395349e-06,
      "loss": 3.6785,
      "step": 160
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.9323529411764706,
      "eval_loss": 3.7415974140167236,
      "eval_runtime": 22.0515,
      "eval_samples_per_second": 46.255,
      "eval_steps_per_second": 5.805,
      "step": 160
    }
  ],
  "max_steps": 192,
  "num_train_epochs": 6,
  "total_flos": 3.955633604923392e+17,
  "trial_name": null,
  "trial_params": null
}