Open
Conversation
Prevent `_prompt_inference_settings()` from raising `UnboundLocalError` when the resolved backend is not `torch` (for example MLX on Apple Silicon). The function always returned `InferenceSettings`, but `generate_comp` and `gpu_post_processing` were only assigned inside the torch-specific branch. On MLX this left both locals undefined and caused inference to crash before startup. Initialize those fields from `InferenceSettings` defaults before the backend check, then continue overriding them only for the torch path. prompt helper returns valid settings instead of crashing.
|
Thanks for the PR, @Raiden129 - my team's running into this problem at the moment and I was just about to dig into a fix myself. Looks like I don't have to. Kudos! |
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Prevent
_prompt_inference_settings()from raisingUnboundLocalErrorwhen the resolved backend is nottorch(for MLX on Apple Silicon).The function always returned
InferenceSettings, butgenerate_compandgpu_post_processingwere only assigned inside the torch-specific branch. On MLX this left both locals undefined and caused inference to crash before startup.Initialize those fields from
InferenceSettingsdefaults before the backend check, then continue overriding them only for the torch path.prompt helper returns valid settings instead of crashing.
What does this change?
Initializes generate_comp and gpu_post_processing before backend-specific prompting.
Preserves existing torch behavior: torch users are still prompted for composition preview and GPU post-processing settings.
Makes non-torch backend MLX fall back to the InferenceSettings defaults instead of crashing.
Checklist
uv run pytestpassesuv run ruff checkpassesuv run ruff format --checkpasses