Make GLM-OCR pip-installable (PyPI package)

[JaimeArboleda]

Feature request / 功能建议

Make GLM-OCR installable via pip.

Suggested Changes:

1. Add a pyproject.toml or setup.py file defining the package.
2. Structure the repository as an installable Python module.
3. Expose a minimal API for programmatic use (e.g., from glm_ocr import OCRModel).
4. Optionally publish releases to PyPI, so that organizations with internal mirrors (like Nexus) can mirror and install easily.

Thanks for the great project!

Motivation / 动机

Currently, installing GLM-OCR requires cloning the repository and manually installing dependencies. Providing a proper Python package would significantly improve usability and adoption.

In our organization, we have access to Python packages only via our internal Nexus repository. Installing directly from Git (e.g., pip install git+https://github.com/zai-org/GLM-OCR.git) is difficult or restricted. Other organizations might face the same limitation.

Being able to install the project with a simple command such as:

pip install glm-ocr

would make it much easier to integrate GLM-OCR into Python workflows, reproducible environments, Docker images, and CI pipelines.

Your contribution / 您的贡献

I have not submitted any PR.

[zRzRzRzRzRzRzR]

Completed