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.

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

Remove reuse_logits_for_grads option for RNNTL #1610

Merged
merged 2 commits into from
Aug 3, 2021

Conversation

carolineechen
Copy link
Contributor

following discussion in #1532 and #1536, we are removing the reuse_logits_for_grad option for RNNT loss, as this results in modifying inputs in place and is not the usual behavior of a loss function

Copy link
Contributor

@vincentqb vincentqb left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

rtol=rtol,
err_msg=f"failed on b={b}, t={t}/T={T}, u={u}/U={U}",
)
with self.subTest():
Copy link
Contributor

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

nit: subTest is no longer needed since there is only one test to run

@carolineechen carolineechen force-pushed the rnntl-remove-reuse-logits branch from 0802145 to 326e34d Compare July 9, 2021 19:29
@carolineechen carolineechen force-pushed the rnntl-remove-reuse-logits branch from 326e34d to d55aed7 Compare August 3, 2021 16:54
@carolineechen carolineechen merged commit 16f3b2f into pytorch:main Aug 3, 2021
mthrok pushed a commit to mthrok/audio that referenced this pull request Dec 13, 2022
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Projects
None yet
Development

Successfully merging this pull request may close these issues.

3 participants