Handle WeightAdapters in ops.move_patch_to_device - #461
Open
blepping wants to merge 1 commit into
Open
Conversation
ChrisColeTech
added a commit
to ChrisColeTech/ComfyUI-GGUF-Loader
that referenced
this pull request
Aug 12, 2026
Cherry-pick production fixes from city96/ComfyUI-GGUF open PRs: - city96#472 dequant device-constant cache (major LTX/sampling speedup) - city96#433 IQ1/IQ2/IQ3 torch dequant (extra TE quants) - city96#470 QK-norm .scale→.weight (silent NaN/black Flux-compat) - city96#467 dequant bare nn.Parameters (LTX learnable_registers) - city96#392 lumina2/zimage pad token shape fix - city96#456/city96#468 GGMLTensor dtype + dequantize() for core cast path - city96#461 WeightAdapter-aware move_patch_to_device - city96#469 force_patch on partial load/unload - city96#440/city96#436 mistral3 TE, city96#438 qwen35, qwen2 allowlist - city96#473 partial: Qwen3-VL deepstack mmproj map for MiniMax H3 TE Skipped mega/draft rewrites (city96#459, full city96#473 LazyGGUFReader, city96#336 Triton). See PR_BACKPORT.md for the full matrix.
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
1 participant
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
move_patch_to_devicehasn't done anything in a really long time since patches are basically never baretorch.Tensorinstances anymore and are instances ofWeightAdapterBaseinstead. The currentmove_patch_to_deviceis a no-op.I am not positive this approach is the correct one or will handle something like an exotic custom instance of the weight adapter base class that behaves differently. I have been using it in my local version for months now without any problems.
model_management.cast_to_deviceautomatically determines of non-blocking transfers are supported and enables that when appropriate.copy=truemight not be necessary, but it seemed like the safe choice. I haven't tried with that disabled.Note: I said this works for me but I have been using it/testing with other local patches like the Triton kernels. I don't see how that could make a difference but like the jerk I am, I haven't actually tested it in a vanilla setup.