test.ipynb fails on fresh setup due to missing pretrained weights and grid artifacts

[kunalsanga]

Description

While running test.ipynb in
Grid_based_strong_lensing_for_unsupervised_super_resolution_Anirudh_Shankar
on a fresh environment, the notebook fails when attempting to load pretrained
artifacts that are not present in the repository.

Specifically, the notebook assumes the following files exist locally:

• vdl_weights.pt
• scatter_to_log_128.pt
• forward_from_log_128.pt
• scatter_from_log_128.pt
• sparse_grid_fracs_euclid_backward.pt

However, the README does not document:

• where these files should be obtained from,
• whether they are pretrained or generated locally,
• or what steps a new user should follow to run test.ipynb end-to-end.

Steps to reproduce

1. Create a fresh Python virtual environment
2. Install dependencies
3. Open and run test.ipynb

Observed behavior

The notebook fails with an error such as:

Expected behavior

A new contributor should either:

• have clear instructions to obtain or generate the required artifacts, or
• see a setup cell/script that downloads them automatically, or
• receive a clear error message explaining the missing prerequisites.

Happy to help implement a fix once the intended workflow is clarified.

[kunalsanga]

For reference, the error raised is:

FileNotFoundError: [Errno 2] No such file or directory: 'vdl_weights.pt'

[iamsarvrath]

Hi @kunalsanga,

Thanks for reporting this issue! I've identified and fixed the problem. Here's what I've done:

Changes Made

1. Added file validation in test.ipynb:

   • Added a setup cell that checks for all required files before attempting to load them
   • Provides clear, actionable error messages when files are missing
   • Fixed the GRID_MATRICES_DIR undefined variable error

2. Updated README.md:

   • Added a new "Running test.ipynb" section with comprehensive documentation
   • Documented all required files and their expected locations
   • Provided step-by-step instructions for:
     • Obtaining vdl_weights.pt (via training or repository releases)
     • Regenerating grid matrices using the appropriate notebooks

Solution Overview

Now when users run test.ipynb on a fresh setup:

• They'll see clear error messages if files are missing
• The error messages include instructions on how to obtain or generate the required files
• The README provides complete documentation for the setup process

The fix ensures that new contributors will have clear guidance instead of encountering cryptic FileNotFoundError messages.

Thanks again for bringing this to attention!