public-mirrors/Kokoro-FastAPI
1
0
Fork 0
mirror of https://github.com/remsky/Kokoro-FastAPI.git synced 2025-08-05 16:48:53 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Fix bug: same reference issue

Browse source
This commit is contained in:
fondoger 2025-04-12 22:08:05 +08:00
parent 5dbf2e2e4b
commit 655c243f1d
2 changed files with 5 additions and 5 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

6
api/src/inference/base.py
View file

@ -12,12 +12,12 @@ class AudioChunk:
def __init__(
self,
audio: np.ndarray,
word_timestamps: Optional[List] = [],
audio: np.ndarray, # dtype: np.int16
word_timestamps: Optional[List] = None, # Using None instead of `[]` to avoid reference to the same empty list.
output: Optional[Union[bytes, np.ndarray]] = b"",
):
self.audio = audio
self.word_timestamps = word_timestamps
self.word_timestamps = word_timestamps if word_timestamps is not None else []
self.output = output
@staticmethod

4
api/src/services/tts_service.py
View file

@ -65,7 +65,7 @@ class TTSService:
# Handle silence tags, eg: `[silent](0.5s)`
if match := SILENCE_TAG.match(chunk_text):
silence_duration = float(match.group(1))
silence_audio = np.zeros(int(silence_duration * 24000), dtype=np.float32)
silence_audio = np.zeros(int(silence_duration * 24000), dtype=np.int16)
if not output_format:
yield AudioChunk(silence_audio, output=b"")
return
@ -89,7 +89,7 @@ class TTSService:
return
chunk_data = await AudioService.convert_audio(
AudioChunk(
np.array([], dtype=np.float32)
np.array([], dtype=np.int16)
), # Dummy data for type checking
output_format,
writer,