Fix streaming dataset features after map/add_column - #8685
Open
dajiaohuang wants to merge 1 commit into
Open
dajiaohuang wants to merge 1 commit into
dajiaohuang wants to merge 1 commit into
Conversation
- Preserve existing features when map() is called without explicit features parameter - Properly remove features for columns specified in remove_columns - Add feature parameter to add_column() and infer new column features automatically - Closes huggingface#5752
Member
|
I feel like it can be too expensive to always have to infer features, that's why it wasn't implement in the first place, wdyt ? |
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Fixes #5752
What changed?
map()is called without an explicitfeaturesparameter (including identity functions likemap(lambda x: x))remove_columnsfeatureparameter toadd_column()(consistent with the non-streamingDataset.add_columnAPI)Why?
Previously, calling
.map()or.add_column()on an IterableDataset (streaming mode) would set features toNone, making it impossible to access dataset features after these operations. This fixes the issue reported in #5752 wheremodified_dataset.features.keys()would raise an AttributeError.