How to run the benchmark Install the dependencies poetry install Export the environment variables source .env Run the classification task poetry run python -m tasks.classification classification --project_id --dataset_id --task_path --task_id