Skip to content

vllm.entrypoints.openai.translations.api_router

logger module-attribute 

logger = init_logger(__name__)

router module-attribute 

router = APIRouter()

attach_router 

attach_router(app: FastAPI)
Source code in vllm/entrypoints/openai/translations/api_router.py 
127
128
def attach_router(app: FastAPI):
    app.include_router(router)

create_transcriptions async 

create_transcriptions(
    raw_request: Request,
    request: Annotated[TranscriptionRequest, Form()],
)
Source code in vllm/entrypoints/openai/translations/api_router.py 
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
@router.post(
    "/v1/audio/transcriptions",
    responses={
        HTTPStatus.OK.value: {"content": {"text/event-stream": {}}},
        HTTPStatus.BAD_REQUEST.value: {"model": ErrorResponse},
        HTTPStatus.UNPROCESSABLE_ENTITY.value: {"model": ErrorResponse},
        HTTPStatus.INTERNAL_SERVER_ERROR.value: {"model": ErrorResponse},
    },
)
@with_cancellation
@load_aware_call
async def create_transcriptions(
    raw_request: Request, request: Annotated[TranscriptionRequest, Form()]
):
    handler = transcription(raw_request)
    if handler is None:
        base_server = raw_request.app.state.openai_serving_tokenization
        return base_server.create_error_response(
            message="The model does not support Transcriptions API"
        )

    audio_data = await request.file.read()
    try:
        generator = await handler.create_transcription(audio_data, request, raw_request)
    except Exception as e:
        return handler.create_error_response(e)
    if isinstance(generator, ErrorResponse):
        return JSONResponse(
            content=generator.model_dump(), status_code=generator.error.code
        )

    elif isinstance(generator, TranscriptionResponseVariant):
        return JSONResponse(content=generator.model_dump())

    return StreamingResponse(content=generator, media_type="text/event-stream")

create_translations async 

create_translations(
    request: Annotated[TranslationRequest, Form()],
    raw_request: Request,
)
Source code in vllm/entrypoints/openai/translations/api_router.py 
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
@router.post(
    "/v1/audio/translations",
    responses={
        HTTPStatus.OK.value: {"content": {"text/event-stream": {}}},
        HTTPStatus.BAD_REQUEST.value: {"model": ErrorResponse},
        HTTPStatus.UNPROCESSABLE_ENTITY.value: {"model": ErrorResponse},
        HTTPStatus.INTERNAL_SERVER_ERROR.value: {"model": ErrorResponse},
    },
)
@with_cancellation
@load_aware_call
async def create_translations(
    request: Annotated[TranslationRequest, Form()], raw_request: Request
):
    handler = translation(raw_request)
    if handler is None:
        base_server = raw_request.app.state.openai_serving_tokenization
        return base_server.create_error_response(
            message="The model does not support Translations API"
        )

    audio_data = await request.file.read()
    try:
        generator = await handler.create_translation(audio_data, request, raw_request)
    except Exception as e:
        return handler.create_error_response(e)

    if isinstance(generator, ErrorResponse):
        return JSONResponse(
            content=generator.model_dump(), status_code=generator.error.code
        )

    elif isinstance(generator, TranslationResponseVariant):
        return JSONResponse(content=generator.model_dump())

    return StreamingResponse(content=generator, media_type="text/event-stream")

init_transcription_state 

init_transcription_state(
    engine_client: EngineClient,
    state: State,
    args: Namespace,
    request_logger: RequestLogger | None,
    supported_tasks: tuple[SupportedTask, ...],
)
Source code in vllm/entrypoints/openai/translations/api_router.py 
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
def init_transcription_state(
    engine_client: "EngineClient",
    state: "State",
    args: "Namespace",
    request_logger: RequestLogger | None,
    supported_tasks: tuple["SupportedTask", ...],
):
    state.openai_serving_transcription = (
        OpenAIServingTranscription(
            engine_client,
            state.openai_serving_models,
            request_logger=request_logger,
            log_error_stack=args.log_error_stack,
            enable_force_include_usage=args.enable_force_include_usage,
        )
        if "transcription" in supported_tasks
        else None
    )
    state.openai_serving_translation = (
        OpenAIServingTranslation(
            engine_client,
            state.openai_serving_models,
            request_logger=request_logger,
            log_error_stack=args.log_error_stack,
            enable_force_include_usage=args.enable_force_include_usage,
        )
        if "transcription" in supported_tasks
        else None
    )

transcription 

transcription(
    request: Request,
) -> OpenAIServingTranscription
Source code in vllm/entrypoints/openai/translations/api_router.py 
44
45
def transcription(request: Request) -> OpenAIServingTranscription:
    return request.app.state.openai_serving_transcription

translation 

translation(request: Request) -> OpenAIServingTranslation
Source code in vllm/entrypoints/openai/translations/api_router.py 
48
49
def translation(request: Request) -> OpenAIServingTranslation:
    return request.app.state.openai_serving_translation