Unable to connect client side script to server side tensorflow_model_server via GRPC

[shantanu-deepquanty]

I'm trying to server my Object Detection model via TF Serving and I'm unable to get results on the client side script.

I'm running this in an environment with tensorflow_model_server version 2.9.1 (built with GPU support) and the pip package tensorflow-serving-api version 2.2.0 All this is being run in a RHEL 8 system with a CUDA version of 11.0

tensorflow_model_server was built by referring to the official TF documentation and with the docker file : https://github.com/tensorflow/serving/blob/master/tensorflow_serving/tools/docker/Dockerfile.gpu

The binary file resulting from the build is being run outside any docker container and on a bare metal server. How do I make tensorflow_model_server listen on IPV4 instead of the IPV6? Is there any way to solve the error shown in the above screenshot?

[singhniraj08]

@shantanu-deepquanty,

Kindly share the minimal code to reproduce the error for faster resolution. In meanwhile, you can refer to documentation and tutorial for using gRPC API with Tensorflow ModelServer. Thank you!

[singhniraj08]

Closing this due to inactivity. Please take a look into the answers provided above, feel free to reopen and post your comments(if you still have queries on this). Thank you!