Prepare processors for VideoLLMs #36149
Merged
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
|
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. |
zucchini-nlp
changed the title
molbap
reviewed
Feb 12, 2025
Contributor
molbap
left a comment
molbap
left a comment
There was a problem hiding this comment.
Thanks for the quick iteration! Added a couple comments of things I'm not sure design or code-wise
Co-authored-by: Pablo Montalvo <39954772+molbap@users.noreply.github.com>
Co-authored-by: Pablo Montalvo <39954772+molbap@users.noreply.github.com>
molbap
approved these changes
Feb 13, 2025
| f"Make sure that fps of a video is less than the requested fps for loading. Detected video_fps={video_fps}" | ||
| ) | ||
| indices = get_uniform_frame_indices(total_num_frames, num_frames=num_frames) | ||
| duration = total_num_frames / video_fps if video_fps else 0 |
Contributor
There was a problem hiding this comment.
random thought, what does it mean to have a duration 0 video?
Member
Author
There was a problem hiding this comment.
An error occured within video decoder and it couldn't give us back the duration. Rarely that can happen
Co-authored-by: Pablo Montalvo <39954772+molbap@users.noreply.github.com>
qubvel
reviewed
Feb 13, 2025
Contributor
qubvel
left a comment
qubvel
left a comment
There was a problem hiding this comment.
Thanks! Made a quick look and left a coment, but it's up to you
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 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Outdated suggestions cannot be applied.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?
As discussed internally, these changes will help SmolVLM2 to be consistent with how processors usually work for other VLMs and lay ground for other video-LLMs in the future. The current API for chat templates should be more flexible now to handle special models
TODO:
At the end we should have the below working before release. We might do a short cut by overriding
apply_chat_templateif some things fail on the way, and then refactor out after releasecc @molbap @yonigozlan