| """ |
| RegFM setup. |
| |
| The customized `transformers` package lives under `src/transformers`. |
| Installing this project (pip install -e .) also installs that package, so that |
| `import transformers` resolves to the local copy. |
| """ |
|
|
| import shutil |
| from pathlib import Path |
|
|
| from setuptools import find_packages, setup |
|
|
| |
| for egg_name in ("transformers.egg-info", "regfm.egg-info"): |
| stale_egg_info = Path(__file__).parent / egg_name |
| if stale_egg_info.exists(): |
| print(f"Warning: {stale_egg_info} exists; removing it for a clean editable install.") |
| shutil.rmtree(stale_egg_info) |
|
|
| extras = { |
| "torch": ["torch"], |
| "train": ["tensorboard"], |
| } |
| extras["all"] = extras["torch"] + extras["train"] |
|
|
| setup( |
| name="regfm", |
| version="0.1.0", |
| author="Zijing Gao", |
| author_email="gzj21@mails.tsinghua.edu.cn", |
| description="RegFM: a regulatory foundation model for gene expression prediction", |
| long_description=open("README.md", "r", encoding="utf-8").read(), |
| long_description_content_type="text/markdown", |
| keywords="genomics gene-expression regulatory deep-learning transformer", |
| url="https://github.com/<your-org>/RegFM", |
| |
| |
| |
| |
| package_dir={"": "src"}, |
| packages=find_packages("src"), |
| py_modules=[ |
| "main", |
| "model", |
| "module", |
| "dataset", |
| "utils", |
| ], |
| install_requires=[ |
| "numpy", |
| "tokenizers==0.15.2", |
| "boto3", |
| "filelock", |
| "requests", |
| "tqdm>=4.27", |
| "regex!=2019.12.17", |
| "sentencepiece", |
| "sacremoses", |
| ], |
| extras_require=extras, |
| python_requires=">=3.11", |
| classifiers=[ |
| "Programming Language :: Python :: 3", |
| "License :: OSI Approved :: MIT License", |
| "Operating System :: OS Independent", |
| "Topic :: Scientific/Engineering :: Bio-Informatics", |
| "Topic :: Scientific/Engineering :: Artificial Intelligence", |
| ], |
| ) |
|
|