trainer_state.json

{
  "best_metric": null,
  "best_model_checkpoint": null,
  "epoch": 3.242424242424242,
  "eval_steps": 25,
  "global_step": 16,
  "is_hyper_param_search": false,
  "is_local_process_zero": true,
  "is_world_process_zero": true,
  "log_history": [
    {
      "epoch": 0.24242424242424243,
      "grad_norm": 10.591867446899414,
      "learning_rate": 0.0001875,
      "loss": 9.5701,
      "step": 1
    },
    {
      "epoch": 0.48484848484848486,
      "grad_norm": Infinity,
      "learning_rate": 0.0001875,
      "loss": 9.4247,
      "step": 2
    },
    {
      "epoch": 0.7272727272727273,
      "grad_norm": 9.98997974395752,
      "learning_rate": 0.000175,
      "loss": 9.4533,
      "step": 3
    },
    {
      "epoch": 0.9696969696969697,
      "grad_norm": 12.13016128540039,
      "learning_rate": 0.00016250000000000002,
      "loss": 9.2022,
      "step": 4
    },
    {
      "epoch": 1.0,
      "grad_norm": 1.9623814821243286,
      "learning_rate": 0.00015000000000000001,
      "loss": 1.1276,
      "step": 5
    },
    {
      "epoch": 1.2424242424242424,
      "grad_norm": 14.68592643737793,
      "learning_rate": 0.0001375,
      "loss": 8.8048,
      "step": 6
    },
    {
      "epoch": 1.4848484848484849,
      "grad_norm": 18.520177841186523,
      "learning_rate": 0.000125,
      "loss": 8.5944,
      "step": 7
    },
    {
      "epoch": 1.7272727272727273,
      "grad_norm": 19.857175827026367,
      "learning_rate": 0.00011250000000000001,
      "loss": 8.2232,
      "step": 8
    },
    {
      "epoch": 1.9696969696969697,
      "grad_norm": 17.700075149536133,
      "learning_rate": 0.0001,
      "loss": 8.0768,
      "step": 9
    },
    {
      "epoch": 2.0,
      "grad_norm": 2.770627498626709,
      "learning_rate": 8.75e-05,
      "loss": 0.9524,
      "step": 10
    },
    {
      "epoch": 2.242424242424242,
      "grad_norm": 11.319780349731445,
      "learning_rate": 7.500000000000001e-05,
      "loss": 7.7231,
      "step": 11
    },
    {
      "epoch": 2.484848484848485,
      "grad_norm": 9.892034530639648,
      "learning_rate": 6.25e-05,
      "loss": 7.6804,
      "step": 12
    },
    {
      "epoch": 2.7272727272727275,
      "grad_norm": 12.017780303955078,
      "learning_rate": 5e-05,
      "loss": 7.663,
      "step": 13
    },
    {
      "epoch": 2.9696969696969697,
      "grad_norm": 12.697382926940918,
      "learning_rate": 3.7500000000000003e-05,
      "loss": 7.5539,
      "step": 14
    },
    {
      "epoch": 3.0,
      "grad_norm": 1.6977847814559937,
      "learning_rate": 2.5e-05,
      "loss": 0.9348,
      "step": 15
    },
    {
      "epoch": 3.242424242424242,
      "grad_norm": 9.705489158630371,
      "learning_rate": 1.25e-05,
      "loss": 7.5541,
      "step": 16
    },
    {
      "epoch": 3.242424242424242,
      "step": 16,
      "total_flos": 2965606484364.0,
      "train_loss": 7.03366831317544,
      "train_runtime": 127.4377,
      "train_samples_per_second": 1.036,
      "train_steps_per_second": 0.126
    }
  ],
  "logging_steps": 1.0,
  "max_steps": 16,
  "num_input_tokens_seen": 0,
  "num_train_epochs": 4,
  "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": 2965606484364.0,
  "train_batch_size": 1,
  "trial_name": null,
  "trial_params": null
}