Repository navigation
Conversation
7 of 10 tasks
chsu-amd
marked this pull request as ready for review
November 19, 2026 04:10
Member
|
Can we reduce the number of PRs being opened / day a little bit? 😅 |
Contributor
Author
|
Sorry about that, will slow down. |
This branch has not been deployed
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
What does this PR do?
Fixes #15347
KolorsPAGPipelineassigned the result ofprepare_ip_adapter_image_embeds()toimage_embedsinstead ofip_adapter_image_embeds. Withip_adapter_image, the PAG loop that follows iteratesip_adapter_image_embeds, which is stillNone, and raisesTypeError. Withip_adapter_image_embeds, the UNet got only the last tensor of that loop instead of the list, which raisesValueErrorwith two IP-Adapters (and goes through the deprecated tensor fallback with one). The block came in with #8934, which took the PAG loop fromStableDiffusionXLPAGPipelinebut kept theimage_embedsvariable fromKolorsPipeline.This PR does what
StableDiffusionXLPAGPipelinedoes: assign toip_adapter_image_embedsand pass that list to the UNet asadded_cond_kwargs["image_embeds"].With the fix, the reproduction script from #15347 prints
OKfor bothip_adapter_imageandip_adapter_image_embeds(onmainit prints theTypeErrorand theValueError).Tests: a
TestKolorsPAGPipelineIPAdapterclass (IPAdapterTesterMixin) intests/pipelines/pag/test_pag_kolors.py; itstest_ip_adapterfails onmainwith theValueErrorabove. The whole file: 28 passed, 17 skipped on CPU, 44 passed, 1 skipped on an RTX 5090.ruff check/ruff format --check(0.9.10) pass on the changed files.Self-review notes (the
self-reviewskill's rubric:review-rules.md,code_style.md,testing.md):IPAdapterTesterMixinpasses precomputed embeds only, so theip_adapter_imagepath (TypeError) is covered by the KolorsPAGPipeline: IP-Adapter inputs raise TypeError / ValueError #15347 reproduction, not by a unit test; same coverage as the other PAG IP-Adapter tests. [Kolors] Fix LoRA loading in KolorsPAGPipeline #15350 also editstest_pag_kolors.py(imports and a new class next to this one), so whichever merges second needs a small rebase.Before submitting
self-reviewskill on the diff?documentation guidelines, and
here are tips on formatting docstrings.
Who can review?
@yiyixuxu @sayakpaul