rchavezj / tch-rs-serving

This project will contain a back end stack with rust web framework Actix for fast http request. Alongside Juniper (Rust + GraphQL) with Rust bindings (tch-rs) for the C++ API of PyTorch (Libtorch).

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

(Under Construction)

tch-rs (PyTorch C++ API Rust Bindings) into production

Build Status Latest version License

+ Tokio (Async) + Tonic (gRPC) + PostGresQL

+ Actix (Web Framework) + Juniper (GraphQL + Rust)

This project will contain a back end stack with rust web framework Actix for fast API. Alongside Juniper, an integration of Rust w/ GraphQL. And Rust bindings (tch-rs) for the C++ API of PyTorch (Libtorch) using Torchscript.

Full credit for sections 00-02
Tokio.rs
Rust fundamentals
Genus-v Programming
Torch Audio Course

I wish to build up my knowledge from the sources referenced to setup a Machine Learning Rust ecosystem into production.

Why Rust for machine learning?

https://www.lpalmieri.com/posts/2019-12-01-taking-ml-to-production-with-rust-a-25x-speedup/

PyTorch production options I want to consider

alt text

Contents:

00. Server From Scratch 01. Todo-Actix
02. Actix with Juniper 03. Libtorch with Actix

About

This project will contain a back end stack with rust web framework Actix for fast http request. Alongside Juniper (Rust + GraphQL) with Rust bindings (tch-rs) for the C++ API of PyTorch (Libtorch).


Languages

Language:Rust 53.3%Language:C 18.6%Language:C++ 18.6%Language:CMake 8.3%Language:Shell 0.6%Language:HTML 0.4%Language:Python 0.2%Language:Dockerfile 0.1%Language:CSS 0.0%