speech_recognition/FunASR
1
0
Fork 0
mirror of https://github.com/modelscope/FunASR synced 2025-09-15 14:48:36 +08:00
Code Issues Actions Packages Projects Releases Wiki Activity

Merge pull request #205 from alibaba-damo-academy/dev_onnx

Browse Source 
export bug fix
This commit is contained in:
zhifu gao 2023-03-10 18:25:29 +08:00 committed by GitHub
commit ef9fb63162
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
3 changed files with 3 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
funasr/export/models/encoder/conformer_encoder.py
View File

@ -61,7 +61,6 @@ class ConformerEncoder(nn.Module):
speech: torch.Tensor,
speech_lengths: torch.Tensor,
):
speech = speech * self._output_size ** 0.5
mask = self.make_pad_mask(speech_lengths)
mask = self.prepare_mask(mask)
if self.embed is None:

1
funasr/export/models/modules/decoder_layer.py
View File

@ -54,6 +54,7 @@ class DecoderLayer(nn.Module):
def forward(self, tgt, tgt_mask, memory, memory_mask, cache=None):
residual = tgt
tgt = self.norm1(tgt)
tgt_q = tgt
tgt_q_mask = tgt_mask
x = residual + self.self_attn(tgt_q, tgt, tgt, tgt_q_mask)

4
funasr/export/models/modules/encoder_layer.py
View File

@ -61,7 +61,7 @@ class EncoderLayerConformer(nn.Module):
if self.feed_forward_macaron is not None:
residual = x
x = self.norm_ff_macaron(x)
x = residual + self.feed_forward_macaron(x)
x = residual + self.feed_forward_macaron(x) * 0.5
residual = x
x = self.norm_mha(x)
@ -81,7 +81,7 @@ class EncoderLayerConformer(nn.Module):
residual = x
x = self.norm_ff(x)
x = residual + self.feed_forward(x)
x = residual + self.feed_forward(x) * 0.5
x = self.norm_final(x)