Inference Metadata on Ultralytics

[PHIKHONAMTHEMBU]

Search before asking

• I have searched the HUB issues and discussions and found no similar questions.

Question

Hi,

I would like to know how can we access metadata of all inferences that has been executed in the ultralytics hub via API, looking for information such as :
Time of API call
Image
Inference results

is such information accessible post API call ?

Additional

No response

[UltralyticsAssistant]

@PHIKHONAMTHEMBU hello,

Thanks for your question regarding the access to inference metadata via the Ultralytics HUB API.

Currently, the Ultralytics HUB doesn't have a dedicated feature integrated for accessing such detailed inference metadata post API call. You won't be able to retrieve exact information like the time of API call, image, and inference results after the API call has been made.

That said, the details of images and inference results are directly available as the response when you call the API for running an inference task. So you can save this information at your client side when you make the API call.

In terms of time of API call, this would be best logged at the time the API request is made.

We appreciate your interest in our platform and your suggestion for this feature. It's feedback like this that helps us improve. Such feature could indeed prove useful in a variety of scenarios and I will convey this feedback to our development team for future updates.

Let me know if there is anything else I can assist you with. Thanks!

[PHIKHONAMTHEMBU]

thank you for your response.

Indeed, in a scenario where we render this as a service to our clients, we do not need to manage separate infrastructure, all there is we give them API documentation and relevant credentials. we then monitor usage, api calls, metrics, ... on the ultralytics side.

[UltralyticsAssistant]

@PHIKHONAMTHEMBU hello,

I understand your viewpoint. Currently, Ultralytics HUB primarily focuses on providing robust and efficient APIs for executing inference tasks and does not inherently support comprehensive operational metrics, logging or usage monitoring services.

However, for tracking and managing API usage and metrics in the interim, I would suggest integrating Ultralytics API usage within your own application or infrastructure monitoring system, utilizing existing capabilities of your tech stack or by employing cloud services specifically designed for API management.

While these might not be the hands-off solutions, they can provide you with the comprehensive tracking and management control you need while using Ultralytics services.

We truly value your feedback and this is an interesting perspective for us to consider as we continue to evolve our platform. I'll ensure your use-case is passed on to our development team for their consideration in future enhancements.

Thank you for using Ultralytics! Should you have more questions or need further clarification, please feel free to ask.