Tensorflow Serving - more names for the same model

Question

I want to redirect two different URLs to the same model request, so that I don't need to create another base_path or version. If I write my config file as bellow, does it cache my model twice?

models.config :

model_config_list {
  config {
    name: 'name1',
    base_path: '/models/model/',
    model_platform: "tensorflow"
  },
  config {
    name: 'name2',
    base_path: '/models/model/',
    model_platform: "tensorflow"
  }
}

Answer 1

So I ran this project to watch over my Docker containers and started modelserving one model. Then I added the same model path and version as I described in the question, but with a different name. The alocated memory for serving was near the double, and with three models it was near three times.