[WIP] Video support in vLLM backend #42919
Open
+231
−125
Conversation
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
zucchini-nlp
changed the title
Contributor
|
[For maintainers] Suggested jobs to run (before merge) run-slow: blip_2, got_ocr2, instructblip, instructblipvideo, internvl, llava_next_video, llava_onevision, perception_lm, qwen2_5_omni, qwen2_5_vl, qwen2_vl, smolvlm, video_llava |
Contributor
|
View the CircleCI Test Summary for this PR: https://huggingface.co/spaces/transformers-community/circle-ci-viz?pr=42919&sha=ce0dfb |
|
The docs for this PR live here. All of your documentation changes will be reflected on that endpoint. The docs are available until 30 days after the last update. |
1 task
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Outdated suggestions cannot be applied.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.
What does this PR do?
Enables vllm-project/vllm#30680 from transformers side. I had this in my local draft for a very long time and some models still don't fit neatly, such as InternVL uses image token-id per video-frame processing or SmolVLM adds timestamps between each frame
I think we will support models that are easy first, in progress for now.
fyi @hmellor