checkpoint-3529/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 2.0,
  "eval_steps": 1000,
  "global_step": 3529,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0005668934240362812,
      "grad_norm": 4.5956525802612305,
      "learning_rate": 0.0,
      "loss": 2.7497,
      "step": 1
    },
    {
      "epoch": 0.5668934240362812,
      "grad_norm": 2.9084548950195312,
      "learning_rate": 2.7045359092818143e-05,
      "loss": 1.9615,
      "step": 1000
    },
    {
      "epoch": 0.5668934240362812,
      "eval_cosine_accuracy": 0.9695753455162048,
      "eval_loss": 1.0098941326141357,
      "eval_runtime": 24.941,
      "eval_samples_per_second": 379.536,
      "eval_steps_per_second": 1.484,
      "step": 1000
    },
    {
      "epoch": 1.1337110481586403,
      "grad_norm": 3.2524373531341553,
      "learning_rate": 2.0745485090298193e-05,
      "loss": 1.7676,
      "step": 2000
    },
    {
      "epoch": 1.1337110481586403,
      "eval_cosine_accuracy": 0.9739066362380981,
      "eval_loss": 0.8953105211257935,
      "eval_runtime": 23.8081,
      "eval_samples_per_second": 397.595,
      "eval_steps_per_second": 1.554,
      "step": 2000
    },
    {
      "epoch": 1.7002832861189803,
      "grad_norm": 2.6741044521331787,
      "learning_rate": 1.4445611087778244e-05,
      "loss": 1.4662,
      "step": 3000
    },
    {
      "epoch": 1.7002832861189803,
      "eval_cosine_accuracy": 0.9747517704963684,
      "eval_loss": 0.8704529404640198,
      "eval_runtime": 23.8746,
      "eval_samples_per_second": 396.488,
      "eval_steps_per_second": 1.55,
      "step": 3000
    }
  ],
  "logging_steps": 1000,
  "max_steps": 5292,
  "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": false
      },
      "attributes": {}
    }
  },
  "total_flos": 0.0,
  "train_batch_size": 256,
  "trial_name": null,
  "trial_params": null
}