Inference speed of the model is slow

[aiswaryauttla]

Hello,

I appreciate the excellent work. I have been using the model for research. I am struggling with high latency when running inference on the model. I am trying to reach production level speed during testing (a good number would be around 1s for prediction on one image). It would be helpful to understand how we can reach high speed when running prediction.

Thank you