[Query] MlFlow server to make conda env in offline mode [v1]

[that-rahul-guy]

Hi all, I am currently using Seldon Code v1 and using the mlflow server v1.16.0 When we pass the s3 artifacts URI, the serving container first creates a conda env and then pip installs all the dependencies from requirements.txt. This requires the machine to have active internet access in order to download the dependencies. Are there any provisions to pass the required libraries via the s3 bucket and install them in the env w/o interenet instead of pulling from the internet?

Point me to appropriate issues/docs if any already. Thanks.