generate_synthetic_dataset.py

import config
from api_wrappers import hf_data_loader
from generation_steps import synthetic_end_to_start


def run():
    df = hf_data_loader.load_processed_rewriting_dataset_as_pandas()

    print(f"End -> start synthesis:")
    print(f"GENERATION_MULTIPLIER = {synthetic_end_to_start.GENERATION_MULTIPLIER}")
    print(f"REL_INSERTIONS_THRESHOLD = {synthetic_end_to_start.REL_INSERTIONS_THRESHOLD}")
    print(f"GENERATION_ATTEMPTS = {synthetic_end_to_start.GENERATION_ATTEMPTS}")
    df = synthetic_end_to_start.transform(df)
    print("Done")

    df.to_csv(config.SYNTHETIC_DATASET_ARTIFACT)


if __name__ == '__main__':
    run()