zhaochenyang20 / prompt2model_test

The training script is in ./train/train_t5_experiment.py.

You can just run python control.py in the same director to launch the training.

The generated datasets are in ./generation/generated_dataset. They are jp2python, normalization, and SQuAD. Note that these three are generated, and each contains roughly 5000 examples. You can check this.

The only thing you need to change is these three lines in the train_t5_experiment.py.

TRAINED_MODEL_ROOT = Path("/home/chenyan3/result/trained_model")
TRAINED_TOKENIZER_ROOT = Path("/home/chenyan3/result/trained_tokenizer")
RESULT_PATH = Path(f"/home/chenyan3/result/{model_store_name}_{task_name}")

Change it to your own root.

if evaluate: means we are doing an evaluation on the generated test set.

if realistic: means we are using the real dataset to test our trained models.

For the Trainer, the parameters are in hyperparameter_choices_keys. You could see all the parameters in the definition of train_model function.