Debug constructing DataFrame from NamedTuple fails without dtype

pytorch / torcharrow

High performance model preprocessing library on PyTorch

https://pytorch.org/torcharrow/beta/index.html

BSD 3-Clause "New" or "Revised" License

649 stars 79 forks source link

Closed jiawein closed 2 years ago

jiawein commented 2 years ago

Summary: Updated the scope.py file. Changed the way to handle the situation when there is data without dtype or columns.

Reviewed By: hanqiwu0704

Differential Revision: D35823382

facebook-github-bot commented 2 years ago

This pull request was exported from Phabricator. Differential Revision: D35823382

facebook-github-bot commented 2 years ago

This pull request was exported from Phabricator. Differential Revision: D35823382