public-mirrors/Kokoro-FastAPI
1
0
Fork 0
mirror of https://github.com/remsky/Kokoro-FastAPI.git synced 2025-04-13 09:39:17 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Combine the language code checks

Browse source
This commit is contained in:
Fireblade 2025-03-09 15:16:45 -04:00
parent f998cf8d01
commit fbdedfb131
1 changed files with 14 additions and 26 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

38
api/src/routers/openai_compatible.py
View file

@ -144,25 +144,13 @@ async def stream_audio_chunks(
if hasattr(request, "return_timestamps"): if hasattr(request, "return_timestamps"):
unique_properties["return_timestamps"]=request.return_timestamps unique_properties["return_timestamps"]=request.return_timestamps
# Determine language code with proper fallback
lang_code = request.lang_code
if not lang_code:
# Use default_voice_code from settings if available
lang_code = settings.default_voice_code
# Otherwise, use first letter of voice name
if not lang_code and voice_name:
lang_code = voice_name[0].lower()
# Log the language code being used
logger.info(f"Starting audio generation with lang_code: {lang_code}")
try: try:
async for chunk_data in tts_service.generate_audio_stream( async for chunk_data in tts_service.generate_audio_stream(
text=request.input, text=request.input,
voice=voice_name, voice=voice_name,
speed=request.speed, speed=request.speed,
output_format=request.response_format, output_format=request.response_format,
lang_code=lang_code, lang_code=request.lang_code,
normalization_options=request.normalization_options, normalization_options=request.normalization_options,
return_timestamps=unique_properties["return_timestamps"], return_timestamps=unique_properties["return_timestamps"],
): ):
@ -215,6 +203,17 @@ async def create_speech(
"pcm": "audio/pcm", "pcm": "audio/pcm",
}.get(request.response_format, f"audio/{request.response_format}") }.get(request.response_format, f"audio/{request.response_format}")
# Determine language code with proper fallback
if not request.lang_code:
# Use default_voice_code from settings if available
request.lang_code = settings.default_voice_code
# Otherwise, use first letter of voice name
if not request.lang_code and voice_name:
request.lang_code = voice_name[0].lower()
# Log the language code being used
logger.info(f"Starting audio generation with lang_code: {request.lang_code}")
# Check if streaming is requested (default for OpenAI client) # Check if streaming is requested (default for OpenAI client)
if request.stream: if request.stream:
# Create generator but don't start it yet # Create generator but don't start it yet
@ -305,24 +304,13 @@ async def create_speech(
else: else:
# Generate complete audio using public interface # Generate complete audio using public interface
# Determine language code with proper fallback
lang_code = request.lang_code
if not lang_code:
# Use default_voice_code from settings if available
lang_code = settings.default_voice_code
# Otherwise, use first letter of voice name
if not lang_code and voice_name:
lang_code = voice_name[0].lower()
# Log the language code being used
logger.info(f"Starting audio generation with lang_code: {lang_code}")
audio_data = await tts_service.generate_audio( audio_data = await tts_service.generate_audio(
text=request.input, text=request.input,
voice=voice_name, voice=voice_name,
speed=request.speed, speed=request.speed,
normalization_options=request.normalization_options, normalization_options=request.normalization_options,
lang_code=lang_code, lang_code=request.lang_code,
) )
audio_data = await AudioService.convert_audio( audio_data = await AudioService.convert_audio(