feat(comfyui): add comfyui image - #11
Merged
Merged
Conversation
ComfyUI v0.34.0 on CUDA-enabled PyTorch, for GPU image and video generation. Built for the RTX 3090 Ti node, so amd64 only - an arm64 copy would be several GB nothing can schedule a GPU onto. torch/torchvision/torchaudio are pinned as a set rather than resolved: upstream wound torchaudio down, so the cu130 index stops at 2.11.0 while torch is already at 2.13.0, and a free resolve downgrades torch to match. Signed-off-by: swibrow <15628653+swibrow@users.noreply.github.com>
Kubernetes args override CMD, so with CMD-only and no ENTRYPOINT a consumer setting args would have the runtime exec the first flag as the binary. Signed-off-by: swibrow <15628653+swibrow@users.noreply.github.com>
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
ComfyUI v0.34.0 on CUDA-enabled PyTorch, for GPU image and video generation workflows. Consumed by
ai/comfyuiin home-ops on the RTX 3090 Ti node, to run Wan 2.2 text-to-video.Notes
2.11.0while torch is already at2.13.0. Letting pip resolve freely downgrades torch to match torchaudio, so all three are explicit and must move together. Renovate trackstorch; a bump there needs the other two checked.requirements.txt, which lists all three unpinned - pip then treats them as satisfied rather than pulling CPU wheels from PyPI over the top.Validation
hadolint -c .hadolint.yamlpasses. The build itself was not run locally - this is an amd64 CUDA image and the dev machine is arm64, so an emulated multi-GB pip install is not a useful check. CI builds it on a native amd64 runner and runs the metadatatest, which is the real gate.