WebAssembly based AI as a Service with Kubernetes

Demo 1: Run locally with WasmEdge

Prerequisite:

• Install Rust with rustup
• Have GCC installed
• Git and cURL

Steps:

1. Install the wasm32-wasi target

rustup target add wasm32-wasi

2. Install wasmedge

curl -sSf https://raw.githubusercontent.com/WasmEdge/WasmEdge/master/utils/install.sh | bash
source $HOME/.wasmedge/env

3. Build the QuickJS interpreter with the WasmEdge Tensorflow extension

cd wasmedge-quickjs
cargo build --target wasm32-wasi --release --features=tensorflow
cd ..

4. Run JS Examples locally

cd js_food
wasmedge-tensorflow-lite --dir .:. wasmedge_quickjs.wasm main.js
cd ..

5. Run Rust examples locally

cd rust_mobilenet
cargo build --target wasm32-wasi --release

We can AOT compile our Rust code to machine native code, and then use WasmEdge sandbox to run the native code.

wasmedgec-tensorflow target/wasm32-wasi/release/classify.wasm classify.so
wasmedge-tensorflow-lite classify.so < grace_hopper.jpg

Demo 2: Deploy as a FAAS

Prerequisite:

• Install Rust with rustup
• Have GCC installed
• Git and cURL
• Vercel CLI

Steps:

1. Install the wasm32-wasi target

rustup target add wasm32-wasi

2. Build the Rust program to WebAssembly bytecode

cd faas
cd api/functions/image-classification/
cargo build --release --target wasm32-wasi

3. Prepare the build artifacts for deployment

cp target/wasm32-wasi/release/classify.wasm ../../

4. Deploy the function

cd ../../../
vercel deploy

Demo 3

Build container

sudo buildah build --annotation "module.wasm.image/variant=compat" -t classify .