trainer_state.json

{
  "best_global_step": 201,
  "best_metric": 0.6686656671664168,
  "best_model_checkpoint": "./results_kcbert_grouped_balanced/checkpoint-201",
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 603,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.24875621890547264,
      "grad_norm": 4.725838661193848,
      "learning_rate": 2.45e-05,
      "loss": 0.6238,
      "step": 50
    },
    {
      "epoch": 0.4975124378109453,
      "grad_norm": 3.734846353530884,
      "learning_rate": 4.9500000000000004e-05,
      "loss": 0.459,
      "step": 100
    },
    {
      "epoch": 0.746268656716418,
      "grad_norm": 3.5573644638061523,
      "learning_rate": 4.5129224652087475e-05,
      "loss": 0.38,
      "step": 150
    },
    {
      "epoch": 0.9950248756218906,
      "grad_norm": 5.902843475341797,
      "learning_rate": 4.015904572564613e-05,
      "loss": 0.3466,
      "step": 200
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.7653927813163482,
      "eval_f1": 0.6686656671664168,
      "eval_loss": 0.5385204553604126,
      "eval_runtime": 7.8505,
      "eval_samples_per_second": 239.986,
      "eval_steps_per_second": 3.821,
      "step": 201
    },
    {
      "epoch": 1.243781094527363,
      "grad_norm": 4.441750526428223,
      "learning_rate": 3.5188866799204776e-05,
      "loss": 0.2598,
      "step": 250
    },
    {
      "epoch": 1.4925373134328357,
      "grad_norm": 4.32944393157959,
      "learning_rate": 3.0218687872763416e-05,
      "loss": 0.2311,
      "step": 300
    },
    {
      "epoch": 1.7412935323383083,
      "grad_norm": 3.940319538116455,
      "learning_rate": 2.524850894632207e-05,
      "loss": 0.2388,
      "step": 350
    },
    {
      "epoch": 1.9900497512437811,
      "grad_norm": 6.620518207550049,
      "learning_rate": 2.0278330019880718e-05,
      "loss": 0.2314,
      "step": 400
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.7494692144373672,
      "eval_f1": 0.6424242424242425,
      "eval_loss": 0.6471742987632751,
      "eval_runtime": 7.5879,
      "eval_samples_per_second": 248.289,
      "eval_steps_per_second": 3.954,
      "step": 402
    },
    {
      "epoch": 2.2388059701492535,
      "grad_norm": 2.4657232761383057,
      "learning_rate": 1.5308151093439365e-05,
      "loss": 0.1475,
      "step": 450
    },
    {
      "epoch": 2.487562189054726,
      "grad_norm": 4.68269157409668,
      "learning_rate": 1.0337972166998012e-05,
      "loss": 0.1217,
      "step": 500
    },
    {
      "epoch": 2.7363184079601988,
      "grad_norm": 6.081990718841553,
      "learning_rate": 5.3677932405566604e-06,
      "loss": 0.1172,
      "step": 550
    },
    {
      "epoch": 2.9850746268656714,
      "grad_norm": 6.272706031799316,
      "learning_rate": 3.9761431411530813e-07,
      "loss": 0.1335,
      "step": 600
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.7611464968152867,
      "eval_f1": 0.6347402597402597,
      "eval_loss": 0.842448890209198,
      "eval_runtime": 7.5784,
      "eval_samples_per_second": 248.601,
      "eval_steps_per_second": 3.959,
      "step": 603
    }
  ],
  "logging_steps": 50,
  "max_steps": 603,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 500,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 1.2477246786125352e+16,
  "train_batch_size": 64,
  "trial_name": null,
  "trial_params": null
}