environment-setup

name	environment-setup
description	Procedures for setting up the environment for research projects involving Python, PyTorch, and NLP models. Use whenever environment requirements (environment.yml) are present.

Environment Setup for NLP Research

When dealing with deep learning research repositories, follow these steps:

1. Review Requirements: Inspect environment.yml or requirements.txt to identify dependencies.
2. Version Checks:
   • python -VV to check the python version.
   • pip freeze to check installed packages.
3. Environment Creation:
   • Use conda env update --file environment.yml or standard pip requirements installation if conda is not available.
4. Validation:
   • Verify that all necessary libraries for training/evaluation are available.
   • Ensure hardware acceleration (CUDA/ROCm) is correctly configured if needed.
   • Log the final environment state.
5. Consistency:
   • Record the environment state to a file (e.g., python_info.txt) to allow for replication.