Request building

RequestData dataclass

Data class for request data.

Attributes:

Name	Type	Description
url	str	The URL of the request.
request_elements	int	The number of request elements.
headers	Optional[Dict[str, str]]	The headers of the request.
parameters	Optional[Dict[str, Union[str, List[str]]]]	The parameters of the request.
data	Optional[Union[str, bytes]]	The data of the request.
payload	Optional[Dict[str, Any]]	The payload of the request.
image_scaling_factors	List[Optional[float]]	The scaling factors of the images.

Source code in inference_sdk/http/utils/request_building.py

@dataclass(frozen=True)
class RequestData:
    """Data class for request data.

    Attributes:
        url: The URL of the request.
        request_elements: The number of request elements.
        headers: The headers of the request.
        parameters: The parameters of the request.
        data: The data of the request.
        payload: The payload of the request.
        image_scaling_factors: The scaling factors of the images.
    """

    url: str
    request_elements: int
    headers: Optional[Dict[str, str]]
    parameters: Optional[Dict[str, Union[str, List[str]]]]
    data: Optional[Union[str, bytes]]
    payload: Optional[Dict[str, Any]]
    image_scaling_factors: List[Optional[float]]

assembly_request_data(url, batch_inference_inputs, headers, parameters, payload, image_placement)

Assemble request data.

Parameters:

Name	Type	Description	Default
url	str	The URL of the request.	required
batch_inference_inputs	List[Tuple[str, Optional[float]]]	The batch inference inputs.	required
headers	Optional[Dict[str, str]]	The headers of the request.	required
parameters	Optional[Dict[str, Union[str, List[str]]]]	The parameters of the request.	required
payload	Optional[Dict[str, Any]]	The payload of the request.	required
image_placement	ImagePlacement	The image placement.	required

Returns:

Type	Description
RequestData	The request data.

Source code in inference_sdk/http/utils/request_building.py

def assembly_request_data(
    url: str,
    batch_inference_inputs: List[Tuple[str, Optional[float]]],
    headers: Optional[Dict[str, str]],
    parameters: Optional[Dict[str, Union[str, List[str]]]],
    payload: Optional[Dict[str, Any]],
    image_placement: ImagePlacement,
) -> RequestData:
    """Assemble request data.

    Args:
        url: The URL of the request.
        batch_inference_inputs: The batch inference inputs.
        headers: The headers of the request.
        parameters: The parameters of the request.
        payload: The payload of the request.
        image_placement: The image placement.

    Returns:
        The request data.
    """
    data = None
    if image_placement is ImagePlacement.DATA and len(batch_inference_inputs) != 1:
        raise ValueError("Only single image can be placed in request `data`")
    if image_placement is ImagePlacement.JSON and payload is None:
        payload = {}
    if image_placement is ImagePlacement.JSON:
        payload = deepcopy(payload)
        payload = inject_images_into_payload(
            payload=payload,
            encoded_images=batch_inference_inputs,
        )
    elif image_placement is ImagePlacement.DATA:
        data = batch_inference_inputs[0][0]
    else:
        raise NotImplemented(
            f"Not implemented request building method for {image_placement}"
        )
    scaling_factors = [e[1] for e in batch_inference_inputs]

    execution_id_value = execution_id.get()
    if execution_id_value:
        headers = headers.copy()
        headers[EXECUTION_ID_HEADER] = execution_id_value

    return RequestData(
        url=url,
        request_elements=len(batch_inference_inputs),
        headers=headers,
        parameters=parameters,
        data=data,
        payload=payload,
        image_scaling_factors=scaling_factors,
    )

prepare_requests_data(url, encoded_inference_inputs, headers, parameters, payload, max_batch_size, image_placement)

Prepare requests data.

Parameters:

Name	Type	Description	Default
url	str	The URL of the request.	required
encoded_inference_inputs	List[Tuple[str, Optional[float]]]	The encoded inference inputs.	required
headers	Optional[Dict[str, str]]	The headers of the request.	required
parameters	Optional[Dict[str, Union[str, List[str]]]]	The parameters of the request.	required
payload	Optional[Dict[str, Any]]	The payload of the request.	required
max_batch_size	int	The maximum batch size.	required
image_placement	ImagePlacement	The image placement.	required

Returns:

Type	Description
List[RequestData]	The list of request data.

Source code in inference_sdk/http/utils/request_building.py

def prepare_requests_data(
    url: str,
    encoded_inference_inputs: List[Tuple[str, Optional[float]]],
    headers: Optional[Dict[str, str]],
    parameters: Optional[Dict[str, Union[str, List[str]]]],
    payload: Optional[Dict[str, Any]],
    max_batch_size: int,
    image_placement: ImagePlacement,
) -> List[RequestData]:
    """Prepare requests data.

    Args:
        url: The URL of the request.
        encoded_inference_inputs: The encoded inference inputs.
        headers: The headers of the request.
        parameters: The parameters of the request.
        payload: The payload of the request.
        max_batch_size: The maximum batch size.
        image_placement: The image placement.

    Returns:
        The list of request data.
    """
    batches = list(
        make_batches(
            iterable=encoded_inference_inputs,
            batch_size=max_batch_size,
        )
    )
    requests_data = []
    for batch_inference_inputs in batches:
        request_data = assembly_request_data(
            url=url,
            batch_inference_inputs=batch_inference_inputs,
            headers=headers,
            parameters=parameters,
            payload=payload,
            image_placement=image_placement,
        )
        requests_data.append(request_data)
    return requests_data