dotnet build dotnet run

HuggingFace Model Download - Samples

This folder contains end-to-end samples demonstrating the unified HuggingFace model download pattern across multiple SDKs:

• C# (.NET 9.0)
• JavaScript/TypeScript (Node.js 18+)
• Rust

All samples show how to:

1. Add a HuggingFace model to the catalog (fast, metadata only)
2. Check if the model is cached
3. Download model files (ONNX, config, etc.)
4. Load the model into memory
5. Run inference
6. Verify model in catalog

---

📁 Sample Projects

• E2E_CSharp_HuggingFace/ — C# end-to-end test, see README
• E2E_JavaScript_HuggingFace/ — TypeScript/JavaScript end-to-end test, see README
• E2E_Rust_HuggingFace/ — Rust end-to-end test (see src/main.rs)

---

🏁 Quick Start

C# E2E Sample

cd samples/E2E_CSharp_HuggingFace
dotnet build
dotnet run

JavaScript/TypeScript E2E Sample

cd samples/E2E_JavaScript_HuggingFace
npm install
npm test

Rust E2E Sample

cd samples/E2E_Rust_HuggingFace
cargo run

---

Prerequisites

• Foundry Local Core must be built and available
• Model cache directory: defaults to ~/.foundry-local
• Network access for HuggingFace API and model download
• Disk space: ~2GB for the test model

See each sample's README for language-specific requirements.

---

🔎 More Info

• See E2E_HuggingFace_README.md for a cross-language overview and test model details.
• Each sample folder contains a README with usage and expected output.