from setuptools import find_packages, setup VERSION = "0.0.1" DESCRIPTION = "Use Partial Satisfiability to tarin RL agent more efficiently guaranteeing that partial satisfaction for a given specification" LONG_DESCRIPTION = """ We give a partial credit for each completion on automaton graph, and propose a logic based reward shaping based on a given automaton graph structure guaranteeing that we can achieve partial satisfiability """ setup( name="psltl", packages=find_packages(), package_data={"psltl": ["py.typed", "version.txt"]}, version=VERSION, install_requires=[ "stable-baselines3==1.7.0", # must be 1.7.0 version of stable baselines!, but because of qrm..crm... "gym==0.21.0", "gymnasium", "cython<3", "logaut", "pythomata", "pickle5", # for save files, and load files ], description="Adaptive Reward Design for Reinforcement Learning in Complex Robotic Tasks", keywords="Adaptive Reward Design for Reinforcement Learning in Complex Robotic Tasks", license="MIT", long_description=LONG_DESCRIPTION, long_description_content_type="text/markdown", python_requires=">=3.6", # because of baseline # PyPI package information. classifiers=[ "Programming Language :: Python :: 3.7", "Programming Language :: Python :: 3.8", ], ) # python setup.py sdist bdist_wheel # twine upload --repository-url https://test.pypi.org/legacy/ dist/* # twine upload dist/*