fix(glm5next): pin multimodal processor revision - #497
Merged
lukealonso merged 1 commit intoAug 28, 2026
Conversation
Pass ModelConfig.revision when the GLM-5.3 native processor loads tokenizer, image, and video metadata. Repository-backed launches therefore keep processor artifacts on the same checkpoint revision as the model weights; local checkpoint behavior is unchanged. Validation: 3 processor tests passed in the source-composed CUDA 13.3 runtime; Ruff check and format validation passed. Assisted-by: OpenAI Codex <codex@openai.com> Signed-off-by: Martin Vit <martin@voipmonitor.org>
|
Warning Review limit reachedNext included review available in 58 minutes. View limit detailsLimit details: You’ve used the included review currently available. You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository. Review configuration: ⚙️ Run configurationConfiguration used: Organization UI Review profile: CHILL Plan: Pro Plus Run ID: 📒 Files selected for processing (2)
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
This was referenced Aug 28, 2026
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Resulting behavior
Status: implemented.
The native GLM-5.3 processing stack passes
ModelConfig.revisionwhen it loads tokenizer, image-processor, and video-processor metadata. Repository-backed launches therefore resolve processor artifacts from the same checkpoint revision as model weights. Local checkpoint paths retain their established behavior.This prevents a pinned model revision from silently combining its weights with processor metadata from a moving repository branch.
Source contract
local-inference-lab/vllm:dev/jovian-judgementatc79f35ca00e8e93e0943a0d79b85b22b18aac939.1bb599d4215aee0c0aa6b7b7534ff98e9ef26cbe.ModelConfig.revision=Nonepreserves Hugging Face default revision resolution.Validation
tests/models/test_glm5next_model.py -k processor: 3 passed, 28 deselected.git diff --check: passed.Duplicate-work check
Open upstream and
local-inference-lab/vllmpull requests were searched for GLM processor, tokenizer, and revision propagation. No open focused pull request passesModelConfig.revisionthrough this GLM native processor path.vllm-project/vllm#41834is a broad DeepSeek V4 and GLM SM120 enablement branch and does not provide this focused revision-propagation contract.Review disclosure
OpenAI Codex assisted with implementation, tests, runtime qualification, and pull-request preparation. Human review of every changed line and the repository-revision contract is required before merge.