checkpoint-846/trainer_state.json

{
  "best_global_step": null,
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.0,
  "eval_steps": 500,
  "global_step": 846,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.17777777777777778,
      "grad_norm": 0.5015225410461426,
      "learning_rate": 4.9000000000000005e-06,
      "loss": 1.9794,
      "step": 50
    },
    {
      "epoch": 0.35555555555555557,
      "grad_norm": 0.3204170763492584,
      "learning_rate": 9.900000000000002e-06,
      "loss": 1.7003,
      "step": 100
    },
    {
      "epoch": 0.5333333333333333,
      "grad_norm": 0.13799017667770386,
      "learning_rate": 1.49e-05,
      "loss": 1.3558,
      "step": 150
    },
    {
      "epoch": 0.7111111111111111,
      "grad_norm": 0.11320392787456512,
      "learning_rate": 1.9900000000000003e-05,
      "loss": 1.1495,
      "step": 200
    },
    {
      "epoch": 0.8888888888888888,
      "grad_norm": 0.09474427998065948,
      "learning_rate": 2.4900000000000002e-05,
      "loss": 0.9722,
      "step": 250
    },
    {
      "epoch": 1.064,
      "grad_norm": 0.14002135396003723,
      "learning_rate": 2.9900000000000002e-05,
      "loss": 0.8676,
      "step": 300
    },
    {
      "epoch": 1.2417777777777779,
      "grad_norm": 0.16521736979484558,
      "learning_rate": 3.49e-05,
      "loss": 0.8131,
      "step": 350
    },
    {
      "epoch": 1.4195555555555557,
      "grad_norm": 0.26585647463798523,
      "learning_rate": 3.99e-05,
      "loss": 0.7925,
      "step": 400
    },
    {
      "epoch": 1.5973333333333333,
      "grad_norm": 0.43122053146362305,
      "learning_rate": 4.49e-05,
      "loss": 0.7773,
      "step": 450
    },
    {
      "epoch": 1.775111111111111,
      "grad_norm": 0.4061725437641144,
      "learning_rate": 4.99e-05,
      "loss": 0.7665,
      "step": 500
    },
    {
      "epoch": 1.775111111111111,
      "eval_loss": 0.7617867588996887,
      "eval_runtime": 91.9001,
      "eval_samples_per_second": 10.881,
      "eval_steps_per_second": 2.72,
      "step": 500
    },
    {
      "epoch": 1.952888888888889,
      "grad_norm": 0.2137136161327362,
      "learning_rate": 4.291907514450868e-05,
      "loss": 0.7596,
      "step": 550
    },
    {
      "epoch": 2.128,
      "grad_norm": 0.2972916066646576,
      "learning_rate": 3.569364161849711e-05,
      "loss": 0.7555,
      "step": 600
    },
    {
      "epoch": 2.3057777777777777,
      "grad_norm": 0.2066715657711029,
      "learning_rate": 2.846820809248555e-05,
      "loss": 0.7522,
      "step": 650
    },
    {
      "epoch": 2.4835555555555557,
      "grad_norm": 0.2598058581352234,
      "learning_rate": 2.124277456647399e-05,
      "loss": 0.7505,
      "step": 700
    },
    {
      "epoch": 2.6613333333333333,
      "grad_norm": 0.1933412402868271,
      "learning_rate": 1.4017341040462428e-05,
      "loss": 0.7499,
      "step": 750
    },
    {
      "epoch": 2.8391111111111114,
      "grad_norm": 0.14771156013011932,
      "learning_rate": 6.791907514450866e-06,
      "loss": 0.7496,
      "step": 800
    }
  ],
  "logging_steps": 50,
  "max_steps": 846,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 3,
  "save_steps": 1000,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.358808545907966e+18,
  "train_batch_size": 2,
  "trial_name": null,
  "trial_params": null
}