deepgeek/CosyVoice
2
0
Fork 0
mirror of https://github.com/FunAudioLLM/CosyVoice.git synced 2026-02-04 17:39:25 +08:00
Code Issues Packages Projects Releases Wiki Activity

add flow decoder tensorrt infer

Browse Source
This commit is contained in:
zhoubofan.zbf
2024-08-29 23:35:07 +08:00
parent 1d881df8b2
commit 5f21aef786
5 changed files with 149 additions and 19 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
cosyvoice/flow/decoder.py
View File

@@ -159,7 +159,7 @@ class ConditionalDecoder(nn.Module):
_type_: _description_
"""
t = self.time_embeddings(t)
t = self.time_embeddings(t).to(t.dtype)
t = self.time_mlp(t)
x = pack([x, mu], "b * t")[0]