trainer_state-gemma-bnb8-QLORA-super_glue-wsc-sequence_classification.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.925925925925926,
  "eval_steps": 1,
  "global_step": 10,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 1.0,
      "grad_norm": 267.1435852050781,
      "learning_rate": 2.5e-05,
      "loss": 3.6544,
      "step": 1
    },
    {
      "epoch": 1.0,
      "eval_accuracy": 0.6212121212121212,
      "eval_loss": 3.440249443054199,
      "eval_runtime": 10.753,
      "eval_samples_per_second": 12.276,
      "eval_steps_per_second": 0.651,
      "step": 1
    },
    {
      "epoch": 1.1851851851851851,
      "grad_norm": 58.776756286621094,
      "learning_rate": 5e-05,
      "loss": 0.7994,
      "step": 2
    },
    {
      "epoch": 1.1851851851851851,
      "eval_accuracy": 0.6212121212121212,
      "eval_loss": 1.934098243713379,
      "eval_runtime": 10.7499,
      "eval_samples_per_second": 12.279,
      "eval_steps_per_second": 0.651,
      "step": 2
    },
    {
      "epoch": 2.0,
      "grad_norm": 195.9375457763672,
      "learning_rate": 4.375e-05,
      "loss": 1.5175,
      "step": 3
    },
    {
      "epoch": 2.0,
      "eval_accuracy": 0.3787878787878788,
      "eval_loss": 2.742687702178955,
      "eval_runtime": 10.7555,
      "eval_samples_per_second": 12.273,
      "eval_steps_per_second": 0.651,
      "step": 3
    },
    {
      "epoch": 2.3703703703703702,
      "grad_norm": 106.63079071044922,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 0.795,
      "step": 4
    },
    {
      "epoch": 2.3703703703703702,
      "eval_accuracy": 0.3787878787878788,
      "eval_loss": 2.8957302570343018,
      "eval_runtime": 10.7308,
      "eval_samples_per_second": 12.301,
      "eval_steps_per_second": 0.652,
      "step": 4
    },
    {
      "epoch": 3.0,
      "grad_norm": 171.42628479003906,
      "learning_rate": 3.125e-05,
      "loss": 1.3457,
      "step": 5
    },
    {
      "epoch": 3.0,
      "eval_accuracy": 0.38636363636363635,
      "eval_loss": 1.5985740423202515,
      "eval_runtime": 10.7661,
      "eval_samples_per_second": 12.261,
      "eval_steps_per_second": 0.65,
      "step": 5
    },
    {
      "epoch": 3.5555555555555554,
      "grad_norm": 127.56062316894531,
      "learning_rate": 2.5e-05,
      "loss": 0.6679,
      "step": 6
    },
    {
      "epoch": 3.5555555555555554,
      "eval_accuracy": 0.5606060606060606,
      "eval_loss": 0.7368663549423218,
      "eval_runtime": 10.7584,
      "eval_samples_per_second": 12.27,
      "eval_steps_per_second": 0.651,
      "step": 6
    },
    {
      "epoch": 4.0,
      "grad_norm": 9.635160446166992,
      "learning_rate": 1.8750000000000002e-05,
      "loss": 0.2683,
      "step": 7
    },
    {
      "epoch": 4.0,
      "eval_accuracy": 0.6212121212121212,
      "eval_loss": 1.010995626449585,
      "eval_runtime": 10.7353,
      "eval_samples_per_second": 12.296,
      "eval_steps_per_second": 0.652,
      "step": 7
    },
    {
      "epoch": 4.7407407407407405,
      "grad_norm": 142.32652282714844,
      "learning_rate": 1.25e-05,
      "loss": 0.7235,
      "step": 8
    },
    {
      "epoch": 4.7407407407407405,
      "eval_accuracy": 0.6212121212121212,
      "eval_loss": 1.204453468322754,
      "eval_runtime": 10.7462,
      "eval_samples_per_second": 12.283,
      "eval_steps_per_second": 0.651,
      "step": 8
    },
    {
      "epoch": 5.0,
      "grad_norm": 59.76286315917969,
      "learning_rate": 6.25e-06,
      "loss": 0.3247,
      "step": 9
    },
    {
      "epoch": 5.0,
      "eval_accuracy": 0.6212121212121212,
      "eval_loss": 1.1849852800369263,
      "eval_runtime": 10.7387,
      "eval_samples_per_second": 12.292,
      "eval_steps_per_second": 0.652,
      "step": 9
    },
    {
      "epoch": 5.925925925925926,
      "grad_norm": 205.0324249267578,
      "learning_rate": 0.0,
      "loss": 1.0798,
      "step": 10
    },
    {
      "epoch": 5.925925925925926,
      "eval_accuracy": 0.6212121212121212,
      "eval_loss": 1.127427577972412,
      "eval_runtime": 10.7462,
      "eval_samples_per_second": 12.283,
      "eval_steps_per_second": 0.651,
      "step": 10
    },
    {
      "epoch": 5.925925925925926,
      "step": 10,
      "total_flos": 1.4890673025245184e+16,
      "train_loss": 1.1176074355840684,
      "train_runtime": 1045.0059,
      "train_samples_per_second": 5.033,
      "train_steps_per_second": 0.01
    }
  ],
  "logging_steps": 1,
  "max_steps": 10,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 10,
  "save_steps": 500,
  "total_flos": 1.4890673025245184e+16,
  "train_batch_size": 20,
  "trial_name": null,
  "trial_params": null
}