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

update func inference_vllm

Browse Source
This commit is contained in:
雾聪
2025-03-01 19:12:13 +08:00
parent 11dbd88947
commit d94908f9d7
1 changed files with 2 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files

5
cosyvoice/llm/llm.py
View File

@@ -536,8 +536,7 @@ class Qwen2LM(TransformerLM):
for request_output in request_outputs: for request_output in request_outputs:
if str(request_output.request_id) != str(request_id): if str(request_output.request_id) != str(request_id):
continue continue
# print(f"request output: {request_output}")
print(f"request output: {request_output}")
out_token = list(request_output.outputs[0].token_ids)[-1] out_token = list(request_output.outputs[0].token_ids)[-1]
if next_fill_index != -1 and len(out_tokens) == next_fill_index: if next_fill_index != -1 and len(out_tokens) == next_fill_index:
top_ids = self.speech_token_size + 2 top_ids = self.speech_token_size + 2
@@ -575,7 +574,7 @@ class Qwen2LM(TransformerLM):
for request_output in request_outputs: for request_output in request_outputs:
if str(request_output.request_id) != str(request_id): if str(request_output.request_id) != str(request_id):
continue continue
print(f"request output: {request_output}") # print(f"request output: {request_output}")
top_ids = list(request_output.outputs[0].token_ids)[-1] top_ids = list(request_output.outputs[0].token_ids)[-1]
out_tokens.append(top_ids) out_tokens.append(top_ids)
if top_ids >= self.speech_token_size: if top_ids >= self.speech_token_size: