What is the installing file size for torch-serve if I want to deploy a model for inference(image, nlp, or graph)?

In fact, I want to use it similar to ONNX javascript that allows me to do inference in the bowser without Restful or backend.
There is a way to run C++, python, or other programming languages in javascript today.

  1. But, what is the size for torch-serve?
  2. Can I run it with script or code rather than Restful API.
  3. And the size for installing the requirement for torch serve?

Torchserve is not meant to run within a browser but would its own service that you can then call with a RESTful endpoint

If you’re looking for an in browser solution then you can take a look at Server Model | Easy to use set of tools to create on-device ML demos on Android and iOS. Unlock the vast potential of AI innovations. (experimental)