vllm.entrypoints.openai.serving_transcription
OpenAIServingTranscription
¶
Bases: OpenAISpeechToText
Handles transcription requests.
Source code in vllm/entrypoints/openai/serving_transcription.py
__init__
¶
__init__(
engine_client: EngineClient,
model_config: ModelConfig,
models: OpenAIServingModels,
*,
request_logger: Optional[RequestLogger],
return_tokens_as_token_ids: bool = False,
)
Source code in vllm/entrypoints/openai/serving_transcription.py
create_transcription
async
¶
create_transcription(
audio_data: bytes,
request: TranscriptionRequest,
raw_request: Request,
) -> Union[
TranscriptionResponse,
AsyncGenerator[str, None],
ErrorResponse,
]
Transcription API similar to OpenAI's API.
See https://platform.openai.com/docs/api-reference/audio/createTranscription for the API specification. This API mimics the OpenAI transcription API.
Source code in vllm/entrypoints/openai/serving_transcription.py
transcription_stream_generator
async
¶
transcription_stream_generator(
request: TranscriptionRequest,
result_generator: list[
AsyncGenerator[RequestOutput, None]
],
request_id: str,
request_metadata: RequestResponseMetadata,
audio_duration_s: float,
) -> AsyncGenerator[str, None]
Source code in vllm/entrypoints/openai/serving_transcription.py
OpenAIServingTranslation
¶
Bases: OpenAISpeechToText
Handles translation requests.
Source code in vllm/entrypoints/openai/serving_transcription.py
__init__
¶
__init__(
engine_client: EngineClient,
model_config: ModelConfig,
models: OpenAIServingModels,
*,
request_logger: Optional[RequestLogger],
return_tokens_as_token_ids: bool = False,
)
Source code in vllm/entrypoints/openai/serving_transcription.py
create_translation
async
¶
create_translation(
audio_data: bytes,
request: TranslationRequest,
raw_request: Request,
) -> Union[
TranslationResponse,
AsyncGenerator[str, None],
ErrorResponse,
]
Translation API similar to OpenAI's API.
See https://platform.openai.com/docs/api-reference/audio/createTranslation for the API specification. This API mimics the OpenAI translation API.
Source code in vllm/entrypoints/openai/serving_transcription.py
translation_stream_generator
async
¶
translation_stream_generator(
request: TranslationRequest,
result_generator: list[
AsyncGenerator[RequestOutput, None]
],
request_id: str,
request_metadata: RequestResponseMetadata,
audio_duration_s: float,
) -> AsyncGenerator[str, None]