fermyon / ai-examples

A collection of serverless apps that show how Fermyon's Serverless AI (currently in private beta) works. Reference: https://developer.fermyon.com/spin/serverless-ai-tutorial

Fermyon Serverless AI Examples

This repository contains examples of using Fermyon Serverless AI.

Prerequisites

The following prerequisites are needed to build and run these examples:

• Spin version 1.5.0 or newer. You can find the install and update instructions here.
• The following plugins:
  • js2wasm 0.6.1 spin plugins install js2wasm.
  • py2wasm 0.3.2 spin plugins install py2wasm.
• Optional, but highly recommended, is to use the Spin Cloud GPU component. This offloads inferencing to Fermyon Cloud GPUs, and thus requires a free account to Fermyon Cloud Serverless AI.
• If you do not want to use the Cloud GPU component, the instructions on how to download the required models for inferencing and embedding can be found here.

Note that Fermyon Cloud support is currently in private beta. To apply for access to the private beta, please fill out this survey.

Examples overview

Here is a table of the following examples:

Example Name	Spin SDK	Inferencing	Embedding	Vector DBs
blog-recommendation-ts	TypeScript		✅	✅
code-generator-rs	Rust	✅
haiku-generator-rs	Rust	✅
haiku-generator-ts	TypeScript	✅
newsfeeder-ts	Typescript	✅
openapi-rs	Rust	✅
sentiment-analysis-py	Python	✅
sentiment-analysis-rs	Rust	✅
sentiment-analysis-ts	Typescript	✅
silly-walk-ts	Typescript	✅

To get started building your own applications, follow these instructions to make sure you have installed the correct version of Spin and the necessary SDKs. If you'd like to learn more about the API, visit the API guide here.