trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 10.015873015873016,
  "eval_steps": 126,
  "global_step": 630,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.0015873015873015873,
      "grad_norm": 1.802138090133667,
      "learning_rate": 1e-05,
      "loss": 2.7056,
      "step": 1
    },
    {
      "epoch": 1.0015873015873016,
      "grad_norm": 1.3845086097717285,
      "learning_rate": 0.00063,
      "loss": 1.2986,
      "step": 63
    },
    {
      "epoch": 2.003174603174603,
      "grad_norm": 0.29827240109443665,
      "learning_rate": 0.0009509433962264152,
      "loss": 0.369,
      "step": 126
    },
    {
      "epoch": 2.003174603174603,
      "eval_loss": 0.2990573048591614,
      "eval_runtime": 76.7071,
      "eval_samples_per_second": 6.479,
      "eval_steps_per_second": 0.821,
      "step": 126
    },
    {
      "epoch": 3.0047619047619047,
      "grad_norm": 0.28911444544792175,
      "learning_rate": 0.0008320754716981132,
      "loss": 0.2589,
      "step": 189
    },
    {
      "epoch": 4.006349206349206,
      "grad_norm": 0.2769618332386017,
      "learning_rate": 0.0007132075471698113,
      "loss": 0.2183,
      "step": 252
    },
    {
      "epoch": 4.006349206349206,
      "eval_loss": 0.24794502556324005,
      "eval_runtime": 75.6398,
      "eval_samples_per_second": 6.571,
      "eval_steps_per_second": 0.833,
      "step": 252
    },
    {
      "epoch": 5.007936507936508,
      "grad_norm": 0.2538459897041321,
      "learning_rate": 0.0005943396226415095,
      "loss": 0.1887,
      "step": 315
    },
    {
      "epoch": 6.0095238095238095,
      "grad_norm": 0.28301894664764404,
      "learning_rate": 0.0004754716981132076,
      "loss": 0.1622,
      "step": 378
    },
    {
      "epoch": 6.0095238095238095,
      "eval_loss": 0.253131628036499,
      "eval_runtime": 75.5929,
      "eval_samples_per_second": 6.575,
      "eval_steps_per_second": 0.833,
      "step": 378
    },
    {
      "epoch": 7.011111111111111,
      "grad_norm": 0.29330751299858093,
      "learning_rate": 0.00035660377358490565,
      "loss": 0.138,
      "step": 441
    },
    {
      "epoch": 8.012698412698413,
      "grad_norm": 0.2697054147720337,
      "learning_rate": 0.0002377358490566038,
      "loss": 0.1124,
      "step": 504
    },
    {
      "epoch": 8.012698412698413,
      "eval_loss": 0.2732747197151184,
      "eval_runtime": 75.8397,
      "eval_samples_per_second": 6.553,
      "eval_steps_per_second": 0.831,
      "step": 504
    },
    {
      "epoch": 9.014285714285714,
      "grad_norm": 0.21484734117984772,
      "learning_rate": 0.0001188679245283019,
      "loss": 0.0883,
      "step": 567
    },
    {
      "epoch": 10.015873015873016,
      "grad_norm": 0.20847243070602417,
      "learning_rate": 0.0,
      "loss": 0.0692,
      "step": 630
    },
    {
      "epoch": 10.015873015873016,
      "eval_loss": 0.2962268590927124,
      "eval_runtime": 75.5859,
      "eval_samples_per_second": 6.575,
      "eval_steps_per_second": 0.833,
      "step": 630
    }
  ],
  "logging_steps": 63,
  "max_steps": 630,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 9223372036854775807,
  "save_steps": 126,
  "stateful_callbacks": {
    "TrainerControl": {
      "args": {
        "should_epoch_stop": false,
        "should_evaluate": false,
        "should_log": false,
        "should_save": true,
        "should_training_stop": true
      },
      "attributes": {}
    }
  },
  "total_flos": 2.1425470562304e+19,
  "train_batch_size": 16,
  "trial_name": null,
  "trial_params": null
}