trainer_state.json

{
  "best_metric": 1.0,
  "best_model_checkpoint": "beit-base-patch16-224-pt22k-ft22k-finetuned-footulcer/checkpoint-57",
  "epoch": 9.696969696969697,
  "eval_steps": 500,
  "global_step": 80,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.97,
      "eval_accuracy": 0.8103448275862069,
      "eval_loss": 0.36133143305778503,
      "eval_runtime": 1.4634,
      "eval_samples_per_second": 79.267,
      "eval_steps_per_second": 2.733,
      "step": 8
    },
    {
      "epoch": 1.21,
      "grad_norm": 8.757495880126953,
      "learning_rate": 4.8611111111111115e-05,
      "loss": 0.5337,
      "step": 10
    },
    {
      "epoch": 1.94,
      "eval_accuracy": 0.9482758620689655,
      "eval_loss": 0.18706370890140533,
      "eval_runtime": 1.1875,
      "eval_samples_per_second": 97.685,
      "eval_steps_per_second": 3.368,
      "step": 16
    },
    {
      "epoch": 2.42,
      "grad_norm": 4.399858474731445,
      "learning_rate": 4.166666666666667e-05,
      "loss": 0.2621,
      "step": 20
    },
    {
      "epoch": 2.91,
      "eval_accuracy": 0.9655172413793104,
      "eval_loss": 0.09208085387945175,
      "eval_runtime": 1.1232,
      "eval_samples_per_second": 103.273,
      "eval_steps_per_second": 3.561,
      "step": 24
    },
    {
      "epoch": 3.64,
      "grad_norm": 7.322921276092529,
      "learning_rate": 3.472222222222222e-05,
      "loss": 0.2071,
      "step": 30
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.9741379310344828,
      "eval_loss": 0.0625535324215889,
      "eval_runtime": 1.1269,
      "eval_samples_per_second": 102.935,
      "eval_steps_per_second": 3.549,
      "step": 33
    },
    {
      "epoch": 4.85,
      "grad_norm": 6.532228469848633,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.1577,
      "step": 40
    },
    {
      "epoch": 4.97,
      "eval_accuracy": 0.9913793103448276,
      "eval_loss": 0.031551431864500046,
      "eval_runtime": 1.1244,
      "eval_samples_per_second": 103.162,
      "eval_steps_per_second": 3.557,
      "step": 41
    },
    {
      "epoch": 5.94,
      "eval_accuracy": 0.9827586206896551,
      "eval_loss": 0.04208489507436752,
      "eval_runtime": 1.1481,
      "eval_samples_per_second": 101.039,
      "eval_steps_per_second": 3.484,
      "step": 49
    },
    {
      "epoch": 6.06,
      "grad_norm": 5.438969612121582,
      "learning_rate": 2.0833333333333336e-05,
      "loss": 0.1296,
      "step": 50
    },
    {
      "epoch": 6.91,
      "eval_accuracy": 1.0,
      "eval_loss": 0.014194484800100327,
      "eval_runtime": 1.1207,
      "eval_samples_per_second": 103.507,
      "eval_steps_per_second": 3.569,
      "step": 57
    },
    {
      "epoch": 7.27,
      "grad_norm": 6.7644944190979,
      "learning_rate": 1.388888888888889e-05,
      "loss": 0.1102,
      "step": 60
    },
    {
      "epoch": 8.0,
      "eval_accuracy": 0.9827586206896551,
      "eval_loss": 0.056951798498630524,
      "eval_runtime": 1.1318,
      "eval_samples_per_second": 102.492,
      "eval_steps_per_second": 3.534,
      "step": 66
    },
    {
      "epoch": 8.48,
      "grad_norm": 7.136204719543457,
      "learning_rate": 6.944444444444445e-06,
      "loss": 0.1344,
      "step": 70
    },
    {
      "epoch": 8.97,
      "eval_accuracy": 1.0,
      "eval_loss": 0.012280683033168316,
      "eval_runtime": 1.1258,
      "eval_samples_per_second": 103.034,
      "eval_steps_per_second": 3.553,
      "step": 74
    },
    {
      "epoch": 9.7,
      "grad_norm": 5.1168999671936035,
      "learning_rate": 0.0,
      "loss": 0.0905,
      "step": 80
    },
    {
      "epoch": 9.7,
      "eval_accuracy": 1.0,
      "eval_loss": 0.01197639387100935,
      "eval_runtime": 1.1257,
      "eval_samples_per_second": 103.049,
      "eval_steps_per_second": 3.553,
      "step": 80
    },
    {
      "epoch": 9.7,
      "step": 80,
      "total_flos": 7.813276079484396e+17,
      "train_loss": 0.20315870493650437,
      "train_runtime": 289.0237,
      "train_samples_per_second": 35.949,
      "train_steps_per_second": 0.277
    }
  ],
  "logging_steps": 10,
  "max_steps": 80,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 7.813276079484396e+17,
  "train_batch_size": 32,
  "trial_name": null,
  "trial_params": null
}