Great works, it helps me a lot.

[Livioni]

Thanks for your work. Is there any plan to support Nvidia Docker and batch inference, i.e., detect two images in one batch?

[Alex-Lekov]

Thank you.

1. To support Nvidia Docker you just need to change the Dockerfile. Unfortunately I don't have a free machine with an Nvidia card right now to check, but you can do it yourself and add the Dockerfile Nvidia to the repository.

2. batch inference - I will think about the implementation, but usually the issue is solved by adding a queue via another microservice like Apache Kafka

[Livioni]

@Alex-Lekov
So thankful for your reply. I will consider your suggestions.
I'm trying to modify the functions to implement the batch operation.