Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Torchscript for RNN Transducer Loss #8

Draft
wants to merge 2 commits into
base: master
Choose a base branch
from
Draft

Torchscript for RNN Transducer Loss #8

wants to merge 2 commits into from

Conversation

vincentqb
Copy link
Owner

@vincentqb vincentqb commented Feb 2, 2021 •
edited
Loading

Add test for torchscriptability of rnnt_loss.

@vincentqb
Copy link
Owner Author

vincentqb commented Feb 2, 2021 •
edited
Loading

Custom autograd functions (e.g. here) are not torchscriptable (see attached tests and output from pytest torchscript_consistency_cpu_test.py -k "test_rnnt_loss").

Relates to pytorch/pytorch#22329 and discussion. One option is to implement the custom operator in C++, here.

@vincentqb
Copy link
Owner Author

vincentqb commented Feb 2, 2021 •
edited
Loading

tutorials: 1 and 2 and in C++

@vincentqb vincentqb force-pushed the transducer_torchscript branch from 000daa3 to aa1b0b8 Compare February 3, 2021 23:28
@vincentqb vincentqb mentioned this pull request Mar 4, 2021
Open
22 tasks
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers
No reviews
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
1 participant
@vincentqb