gme qwen vl runners fix#15899
Conversation
Summary of ChangesHello @KnightLTC, I'm Gemini Code Assist1! I'm currently reviewing this pull request and will post my feedback shortly. In the meantime, here's a summary to help you and other reviewers quickly get up to speed! This pull request implements a critical fix for the Qwen-VL model runner by updating the mechanism for obtaining input embeddings. This adjustment ensures that the model correctly processes its inputs, resolving a potential compatibility or functional issue and contributing to the overall stability and accuracy of the Qwen-VL integration. Highlights
🧠 New Feature in Public Preview: You can now enable Memory to help Gemini Code Assist learn from your team's feedback. This makes future code reviews more consistent and personalized to your project's style. Click here to enable Memory in your admin console. Using Gemini Code AssistThe full guide for Gemini Code Assist can be found on our documentation page, here are some quick tips. Invoking Gemini You can request assistance from Gemini at any point by creating a comment using either
Customization To customize Gemini Code Assist for GitHub experience, repository maintainers can create a configuration file and/or provide a custom code review style guide (such as PEP-8 for Python) by creating and adding files to a Limitations & Feedback Gemini Code Assist may make mistakes. Please leave feedback on any instances where its feedback is incorrect or counter productive. You can react with 👍 and 👎 on @gemini-code-assist comments. If you're interested in giving your feedback about your experience with Gemini Code Assist for Github and other Google products, sign up here. You can also get AI-powered code generation, chat, as well as code reviews directly in the IDE at no cost with the Gemini Code Assist IDE Extension. Footnotes
|
There was a problem hiding this comment.
Code Review
This pull request fixes a bug in the _forward_gme_qwen2_vl function within the HFRunner. The original code was trying to access a non-existent embed_tokens attribute, which would cause an AttributeError. The fix correctly uses get_input_embeddings() to retrieve the embedding layer. I've suggested a minor improvement to use the higher-level public API for getting the embeddings, which makes the code slightly cleaner and more robust to future changes in the model's internal structure.
| ) -> torch.Tensor: | ||
| if inputs_embeds is None: | ||
| inputs_embeds = self.model.model.embed_tokens(input_ids) | ||
| inputs_embeds = self.model.model.get_input_embeddings()(input_ids) |
There was a problem hiding this comment.
While this change correctly fixes the AttributeError, it's generally better to use the public API of the model class. You could simplify this by using self.model.get_input_embeddings() which is the standard way to get the input embedding layer for Hugging Face models and is equivalent to the current implementation.
| inputs_embeds = self.model.model.get_input_embeddings()(input_ids) | |
| inputs_embeds = self.model.get_input_embeddings()(input_ids) |
…glang into eagle-sche * 'ifmn/eagle-dp-attn' of https://github.com/sgl-project/sglang: (22 commits) dp scheduler enhance support with chunked prefill (sgl-project#16071) modify suffix decoding CI dependency update (sgl-project#16063) fix rotary_embedding init npu (sgl-project#16011) feat: bugfix and accuracy fix for stablelm2_1_6b (sgl-project#15932) Update model and feature support for Ascend NPU (sgl-project#16005) Bugfix for Llama4 (sgl-project#15929) Bugfix for ds-vl2 (sgl-project#15894) gme qwen vl runners fix (sgl-project#15899) add profiling in scheduler (sgl-project#15876) llama use triton rope op (sgl-project#15855) suffix decoding adapt npu suffix decoding adapt npu Add suffix decoding speculative algorithm from feature 13553 cherry sgl-project#15434: qwen3 vl performance update cherry sgl-project#15597: fix Qwen3-VL-30B-A3B-Instruct accuracy loss [Schedule] bug fix for schedule enhancer (sgl-project#15834) minilb support roundrobin (sgl-project#15824) fix torchair compile issue cherry sgl-project#15187: lora fix ... # Conflicts: # python/sglang/srt/managers/scheduler.py # python/sglang/srt/managers/scheduler_enhancer.py
Motivation
Modifications
Accuracy Tests
Benchmarking and Profiling
Checklist