lllyasviel / lllyasviel/stable-diffusion-webui-forge

test_for_nans wiped out in Stable Forge, worth restoring?

Open
#2,261 2 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
13k
Forks
1.7k
PR merge metrics
No merged PRs in 30d

Description

Migrated a utility from Automatic1111 API to Stable Forge API and noticed that existing logic to detect failed all-black image results continues to work in Automatic1111 but not Stable Forge.

Digging into the code it's because the test_for_nans check is just returning in Stable Forge:
https://github.com/lllyasviel/stable-diffusion-webui-forge/blob/4d75507c3d6687439af17d342234bd885be71e03/modules/devices.py#L97
vs
Raising Nans exception in Automatic:
https://github.com/AUTOMATIC1111/stable-diffusion-webui/blob/82a973c04367123ae98bd9abdf80d9eda9b910e2/modules/devices.py#L242

Was this wiped out for a good reason? I can submit a pull request restoring this useful functionality but wanted to verify it's something that is still desired and not prevented by other functionality changes?

Contributor guide

No contributing guide indexed for this repository

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start in modules/devices.py at the cited test_for_nans implementation, then compare it with the referenced Automatic1111 implementation. Determine why the check returns in Stable Forge and whether restoring the exception behavior is compatible with the surrounding functionality. Done means reaching a maintainer-backed decision and, if restoration is wanted, verifying the failed all-black image case.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
machine-learning
Issue type
Bug
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Mostly clear
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.