deepgeek/CosyVoice
2
0
Fork 0
mirror of https://github.com/FunAudioLLM/CosyVoice.git synced 2026-02-05 18:09:24 +08:00
Code Issues Packages Projects Releases Wiki Activity

fix lint

Browse Source
This commit is contained in:
lyuxiang.lx
2024-12-16 09:54:24 +08:00
parent 6b5931dc70
commit ac70560364
8 changed files with 33 additions and 32 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
cosyvoice/cli/cosyvoice.py
View File

@@ -120,6 +120,7 @@ class CosyVoice:
yield model_output
start_time = time.time()
class CosyVoice2(CosyVoice):
def __init__(self, model_dir, load_jit=False, load_onnx=False, load_trt=False):
@@ -153,4 +154,4 @@ class CosyVoice2(CosyVoice):
self.model.load_onnx('{}/flow.decoder.estimator.fp32.onnx'.format(model_dir))
if load_trt:
self.model.load_trt('{}/flow.decoder.estimator.fp16.Volta.plan'.format(model_dir))
del configs
del configs

4
cosyvoice/cli/frontend.py
View File

@@ -123,7 +123,7 @@ class CosyVoiceFrontEnd:
text = remove_bracket(text)
text = re.sub(r'[，,、]+$', '。', text)
texts = list(split_paragraph(text, partial(self.tokenizer.encode, allowed_special=self.allowed_special), "zh", token_max_n=80,
token_min_n=60, merge_len=20, comma_split=False))
token_min_n=60, merge_len=20, comma_split=False))
else:
if self.use_ttsfrd:
texts = [i["text"] for i in json.loads(self.frd.do_voicegen_frd(text))["sentences"]]
@@ -132,7 +132,7 @@ class CosyVoiceFrontEnd:
text = self.en_tn_model.normalize(text)
text = spell_out_number(text, self.inflect_parser)
texts = list(split_paragraph(text, partial(self.tokenizer.encode, allowed_special=self.allowed_special), "en", token_max_n=80,
token_min_n=60, merge_len=20, comma_split=False))
token_min_n=60, merge_len=20, comma_split=False))
if split is False:
return text
return texts

16
cosyvoice/cli/model.py
View File

@@ -330,13 +330,13 @@ class CosyVoice2Model:
def token2wav(self, token, prompt_token, prompt_feat, embedding, uuid, token_offset, finalize=False, speed=1.0):
tts_mel, _ = self.flow.inference(token=token.to(self.device),
token_len=torch.tensor([token.shape[1]], dtype=torch.int32).to(self.device),
prompt_token=prompt_token.to(self.device),
prompt_token_len=torch.tensor([prompt_token.shape[1]], dtype=torch.int32).to(self.device),
prompt_feat=prompt_feat.to(self.device),
prompt_feat_len=torch.tensor([prompt_feat.shape[1]], dtype=torch.int32).to(self.device),
embedding=embedding.to(self.device),
finalize=finalize)
token_len=torch.tensor([token.shape[1]], dtype=torch.int32).to(self.device),
prompt_token=prompt_token.to(self.device),
prompt_token_len=torch.tensor([prompt_token.shape[1]], dtype=torch.int32).to(self.device),
prompt_feat=prompt_feat.to(self.device),
prompt_feat_len=torch.tensor([prompt_feat.shape[1]], dtype=torch.int32).to(self.device),
embedding=embedding.to(self.device),
finalize=finalize)
tts_mel = tts_mel[:, :, token_offset * self.flow.token_mel_ratio:]
# append hift cache
if self.hift_cache_dict[uuid] is not None:
@@ -418,4 +418,4 @@ class CosyVoice2Model:
yield {'tts_speech': this_tts_speech.cpu()}
with self.lock:
self.tts_speech_token_dict.pop(this_uuid)
self.llm_end_dict.pop(this_uuid)
self.llm_end_dict.pop(this_uuid)