1 - GarpePipe using around 7GB of RAM memory after the first request (model size is 500mb)
2- I didn't saw any function for multiple models server. For multi-models, running different servers on different port leads to a lot of RAM memory allocation.
1 - GarpePipe using around 7GB of RAM memory after the first request (model size is 500mb) 2- I didn't saw any function for multiple models server. For multi-models, running different servers on different port leads to a lot of RAM memory allocation.