push_to_hub() for videos - #7971
Merged
Merged
Conversation
|
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. |
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
possible now that row group sizes are auto-determined based on the content size after #7589
Videos are uploaded as PLAIN in Parquet to make sure they can be seeked remotely and with random access to frames in #7976
In the future it could be cool to have the same behavior as when videos are separate files, i.e. lazily load them instead of downloading them completely in streaming mode.
Right now there is this discrepency:
load_dataset("username/my-folder-of-videos", streaming=True)-> videos are lazy loaded one by one when iterating, and only actually downloaded when accessing frames intorchcodecload_dataset("username/my-video-dataset-in-parquet", streaming=True)-> videos are downloaded one by one when iterating, even if no frame is accessed intorchcodecclose #7493