-
Notifications
You must be signed in to change notification settings - Fork 1.9k
[Hackathon 7th] fix aishell3/vctk vc0/ernie #3928
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
Conversation
Thanks for your contribution! |
@@ -1115,7 +1115,8 @@ def forward( | |||
paddle.reshape(xs_pad, (-1, self.odim))), | |||
axis=-1) | |||
mlm_loss = paddle.sum((loss * paddle.reshape( | |||
mlm_loss_pos, [-1]))) / paddle.sum((mlm_loss_pos) + 1e-10) | |||
mlm_loss_pos, | |||
[-1]).astype(loss.dtype))) / paddle.sum((mlm_loss_pos) + 1e-10) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
what's the data format of loss and mlm_loss_pos? paddle.sum((mlm_loss_pos) + 1e-10)
don't need a .astype
?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
mlm_loss_pos 大概是个 bool
@@ -171,7 +171,9 @@ def forward( | |||
if paddle.sum(att_prev) == 0: | |||
# if no bias, 0 0-pad goes 0 | |||
att_prev = 1.0 - make_pad_mask(enc_hs_len) | |||
att_prev = att_prev / enc_hs_len.unsqueeze(-1) | |||
unsqueeze_enc_hs_len = enc_hs_len.unsqueeze(-1).astype( |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
why add new line here?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM
PR types
Others
PR changes
Others
Describe
Others