Lightrun Answers was designed to reduce the constant googling that comes with debugging 3rd party libraries. It collects links to all the places you might be looking at while hunting down a tough bug.
And, if you’re still stuck at the end, we’re happy to hop on a call to see how we can help out.
skip_blank has different values during data preparation
See original GitHub issueI’m running a standard training with parallel sentences containing empty source or target lines and data_io.py returns an error while building buckets:
IndexError: index 897268 is out of bounds for axis 0 with size 897268
The parallel_iter() function in data_io.py is always called with the skip_blank argument set as True, except right here. This line makes us keep the sentence pairs containing “blanks”, which seems to bring the mismatch reflected in the error above. I don’t get the error anymore when I set skip_blank to True (or when I remove the sentence pairs containing blanks in the data).
@mjpost This line came with this PR. Would it be an issue to let the default skip_blank value here ?
Issue Analytics
- State:
- Created 5 years ago
- Comments:6 (6 by maintainers)
Top Results From Across the Web
Top Related Medium Post
Top Related StackOverflow Question
Troubleshoot Live Code
Top Related Reddit Thread
Top Related Hackernoon Post
Top Related Tweet
Top Related Dev.to Post
Top Related Hashnode Post
Thanks for reporting! This indeed doesn’t seem right.
I should also have written, thanks for tracking this down and filing a perfect bug report. It seems you covered a real hole in my use cases 😃