trainer_state.json

{
  "best_metric": 0.9955555555555555,
  "best_model_checkpoint": "train_binary/05-13-2024_12:03:22/checkpoint-50",
  "epoch": 2.5,
  "eval_steps": 10,
  "global_step": 50,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.5,
      "grad_norm": 19.808202743530273,
      "learning_rate": 5e-05,
      "loss": 0.6666,
      "step": 10
    },
    {
      "epoch": 0.5,
      "eval_acc_product": 0.7189312177835052,
      "eval_loss": 0.4033311605453491,
      "eval_overall_acc": 0.84,
      "eval_runtime": 2.1261,
      "eval_samples_per_second": 211.659,
      "eval_steps_per_second": 1.881,
      "eval_voice_active_acc": 0.9948453608247423,
      "eval_voice_passive_acc": 0.72265625,
      "step": 10
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.655267357826233,
      "learning_rate": 4.4444444444444447e-05,
      "loss": 0.2138,
      "step": 20
    },
    {
      "epoch": 1.0,
      "eval_acc_product": 0.9652867268041238,
      "eval_loss": 0.05905359983444214,
      "eval_overall_acc": 0.9844444444444445,
      "eval_runtime": 2.142,
      "eval_samples_per_second": 210.082,
      "eval_steps_per_second": 1.867,
      "eval_voice_active_acc": 0.9690721649484536,
      "eval_voice_passive_acc": 0.99609375,
      "step": 20
    },
    {
      "epoch": 1.5,
      "grad_norm": 4.259519577026367,
      "learning_rate": 3.888888888888889e-05,
      "loss": 0.0748,
      "step": 30
    },
    {
      "epoch": 1.5,
      "eval_acc_product": 0.9652867268041238,
      "eval_loss": 0.027125049382448196,
      "eval_overall_acc": 0.9844444444444445,
      "eval_runtime": 2.1321,
      "eval_samples_per_second": 211.057,
      "eval_steps_per_second": 1.876,
      "eval_voice_active_acc": 0.9690721649484536,
      "eval_voice_passive_acc": 0.99609375,
      "step": 30
    },
    {
      "epoch": 2.0,
      "grad_norm": 0.7526319622993469,
      "learning_rate": 3.3333333333333335e-05,
      "loss": 0.0464,
      "step": 40
    },
    {
      "epoch": 2.0,
      "eval_acc_product": 0.9806902384020619,
      "eval_loss": 0.02873048558831215,
      "eval_overall_acc": 0.9911111111111112,
      "eval_runtime": 2.1428,
      "eval_samples_per_second": 210.002,
      "eval_steps_per_second": 1.867,
      "eval_voice_active_acc": 0.9845360824742269,
      "eval_voice_passive_acc": 0.99609375,
      "step": 40
    },
    {
      "epoch": 2.5,
      "grad_norm": 2.319009780883789,
      "learning_rate": 2.777777777777778e-05,
      "loss": 0.044,
      "step": 50
    },
    {
      "epoch": 2.5,
      "eval_acc_product": 0.9909592461340206,
      "eval_loss": 0.02411176823079586,
      "eval_overall_acc": 0.9955555555555555,
      "eval_runtime": 2.1361,
      "eval_samples_per_second": 210.661,
      "eval_steps_per_second": 1.873,
      "eval_voice_active_acc": 0.9948453608247423,
      "eval_voice_passive_acc": 0.99609375,
      "step": 50
    }
  ],
  "logging_steps": 10,
  "max_steps": 100,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "save_steps": 10,
  "total_flos": 1486430524508160.0,
  "train_batch_size": 128,
  "trial_name": null,
  "trial_params": null
}