Acceleration for consecutive API calling

[FaithFlow]

Hi, thanks for the powerful model. I'm consecutively calling the API to extract objects in many local images. But the speed is slow. It takes about 5-6s per image. Is there any method to accelerate this process? Thank you.

[xifanii]

@FaithFlow Thanks for the feedback!
If you’re using the dds-cloudapi-sdk for API calls, we recommend creating the task with local image auto-resize to speed up file upload and processing. like：

# 2.1. Create task with local image auto resized (Recommended: faster processing)
from dds_cloudapi_sdk import create_task_with_local_image_auto_resize

task = create_task_with_local_image_auto_resize(
    api_path=api_path,
    api_body_without_image=api_body,
    image_path="local/path/to/infer/image.jpg",
)

[FaithFlow]

@xifanii It reduces to around 4s per image, which is still slow. But thank you.