[Feature] Conda Environment Support for Easier and Stable Setup

[Boburmirzo]

Is your feature request related to a problem?

This issue proposes by @BaoZhuhan adding an environment.yml file to the repository to streamline the development setup process for contributors.

Some project dependencies such as faiss-cpu, grpcio, and psycopg involve complex C/C++ extensions. Installing them with pip can sometimes lead to build failures depending on the operating system or required system libraries.

Using Conda for these heavy binary packages provides precompiled builds, making installation more stable, faster, and more predictable across environments

This can significantly improve onboarding for new contributors.

The Solution

The proposed environment.yml file is based on the current pyproject.toml and follows a hybrid structure:

Conda (conda-forge)

Used for binary-heavy dependencies to ensure installation reliability:

• faiss-cpu
• numpy
• grpcio
• psycopg
• other compiled libs

Before we proceed, we need at least 3 developers supporting or requesting this enhancement per our feature request policy.

If you would benefit from Conda-based setup or have suggestions, please leave a comment or 👍 reaction.

Alternatives Considered

No response

Affected Components

Other

Participation

• I am willing to submit a pull request for this feature.