trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 0.9987368421052631,
  "eval_steps": 100,
  "global_step": 593,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.17,
      "grad_norm": 0.4642287790775299,
      "learning_rate": 1.972335510674149e-05,
      "loss": 1.7213,
      "step": 100
    },
    {
      "epoch": 0.17,
      "eval_loss": 1.06580650806427,
      "eval_runtime": 102.5963,
      "eval_samples_per_second": 4.873,
      "eval_steps_per_second": 1.218,
      "step": 100
    },
    {
      "epoch": 0.34,
      "grad_norm": 0.873336136341095,
      "learning_rate": 1.678421935868953e-05,
      "loss": 1.0726,
      "step": 200
    },
    {
      "epoch": 0.34,
      "eval_loss": 1.022551417350769,
      "eval_runtime": 102.2722,
      "eval_samples_per_second": 4.889,
      "eval_steps_per_second": 1.222,
      "step": 200
    },
    {
      "epoch": 0.51,
      "grad_norm": 0.7002689838409424,
      "learning_rate": 1.1555611635532276e-05,
      "loss": 1.0456,
      "step": 300
    },
    {
      "epoch": 0.51,
      "eval_loss": 1.0069515705108643,
      "eval_runtime": 102.5086,
      "eval_samples_per_second": 4.878,
      "eval_steps_per_second": 1.219,
      "step": 300
    },
    {
      "epoch": 0.67,
      "grad_norm": 0.7483100295066833,
      "learning_rate": 5.8020313170922526e-06,
      "loss": 1.0417,
      "step": 400
    },
    {
      "epoch": 0.67,
      "eval_loss": 0.9990952014923096,
      "eval_runtime": 102.3159,
      "eval_samples_per_second": 4.887,
      "eval_steps_per_second": 1.222,
      "step": 400
    },
    {
      "epoch": 0.84,
      "grad_norm": 0.563865602016449,
      "learning_rate": 1.46514039852793e-06,
      "loss": 1.0169,
      "step": 500
    },
    {
      "epoch": 0.84,
      "eval_loss": 0.9966371059417725,
      "eval_runtime": 102.2788,
      "eval_samples_per_second": 4.889,
      "eval_steps_per_second": 1.222,
      "step": 500
    }
  ],
  "logging_steps": 100,
  "max_steps": 593,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 1,
  "save_steps": 500,
  "total_flos": 1.001225674624598e+18,
  "train_batch_size": 4,
  "trial_name": null,
  "trial_params": null
}