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

update vc/tts code

Browse Source
This commit is contained in:
lyuxiang.lx
2024-09-26 11:53:10 +08:00
parent 49015f63e6
commit 72b89a52fb
5 changed files with 74 additions and 21 deletions
Download Patch File Download Diff File Expand all files Collapse all files

2
cosyvoice/flow/flow.py
View File

@@ -125,7 +125,7 @@ class MaskedDiffWithXvec(torch.nn.Module):
h, h_lengths = self.encoder(token, token_len)
h = self.encoder_proj(h)
mel_len1, mel_len2 = prompt_feat.shape[1], int(token_len2 / self.input_frame_rate * 22050 / 256)
h, h_lengths = self.length_regulator.inference(h[:, :token_len1], h[:, token_len1:], mel_len1, mel_len2)
h, h_lengths = self.length_regulator.inference(h[:, :token_len1], h[:, token_len1:], mel_len1, mel_len2, self.input_frame_rate)
# get conditions
conds = torch.zeros([1, mel_len1 + mel_len2, self.output_size], device=token.device)