last-checkpoint/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 5.0,
  "eval_steps": 500,
  "global_step": 5695,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.000877963125548727,
      "grad_norm": 9.281822204589844,
      "learning_rate": 0.0,
      "loss": 5.8495,
      "step": 1
    },
    {
      "epoch": 1.0,
      "grad_norm": 8.733197212219238,
      "learning_rate": 5e-05,
      "loss": 3.0136,
      "step": 1139
    },
    {
      "epoch": 1.0,
      "eval_cosine_accuracy": 0.9113098382949829,
      "eval_loss": 0.8482398986816406,
      "eval_runtime": 34.6129,
      "eval_samples_per_second": 274.608,
      "eval_steps_per_second": 1.098,
      "step": 1139
    },
    {
      "epoch": 2.0,
      "grad_norm": 8.863326072692871,
      "learning_rate": 2.5010969723562967e-05,
      "loss": 2.2096,
      "step": 2278
    },
    {
      "epoch": 2.0,
      "eval_cosine_accuracy": 0.9241451621055603,
      "eval_loss": 0.7464810013771057,
      "eval_runtime": 34.2459,
      "eval_samples_per_second": 277.551,
      "eval_steps_per_second": 1.11,
      "step": 2278
    },
    {
      "epoch": 3.0,
      "grad_norm": 10.734906196594238,
      "learning_rate": 2.1939447125932427e-08,
      "loss": 1.966,
      "step": 3417
    },
    {
      "epoch": 3.0,
      "eval_cosine_accuracy": 0.9337190985679626,
      "eval_loss": 0.6979679465293884,
      "eval_runtime": 34.569,
      "eval_samples_per_second": 274.957,
      "eval_steps_per_second": 1.099,
      "step": 3417
    },
    {
      "epoch": 4.0,
      "grad_norm": 9.933394432067871,
      "learning_rate": 1.251920122887865e-05,
      "loss": 1.8731,
      "step": 4556
    },
    {
      "epoch": 4.0,
      "eval_cosine_accuracy": 0.9330878257751465,
      "eval_loss": 0.700294554233551,
      "eval_runtime": 33.7886,
      "eval_samples_per_second": 281.308,
      "eval_steps_per_second": 1.125,
      "step": 4556
    },
    {
      "epoch": 5.0,
      "grad_norm": 8.708077430725098,
      "learning_rate": 2.1944261575597982e-08,
      "loss": 1.7998,
      "step": 5695
    },
    {
      "epoch": 5.0,
      "eval_cosine_accuracy": 0.9403471946716309,
      "eval_loss": 0.6516265273094177,
      "eval_runtime": 33.9685,
      "eval_samples_per_second": 279.818,
      "eval_steps_per_second": 1.119,
      "step": 5695
    }
  ],
  "logging_steps": 500,
  "max_steps": 5695,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 5,
  "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": 0.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}