-
Notifications
You must be signed in to change notification settings - Fork 140
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
quetion about the loss and grad of "mbr" #322
Comments
@Cescfangs yes |
Thanks for the reply, and I'm curious about the improvement of this mWER tuning, say 5% relative wer reduction? |
neural_sp/neural_sp/models/criterion.py Lines 12 to 39 in 2b10b9c
Also, I am a little confused about the “mbr” loss, the inputs are not used in backward function, how does the grad flow to model params? |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
neural_sp/neural_sp/models/seq2seq/decoders/las.py
Lines 535 to 548 in 2b10b9c
I don't know much about mbr, according to these lines, it looks like a mWER loss and gradient to me
The text was updated successfully, but these errors were encountered: