zk-neural-rust-core

This project provides a Rust core library for zero-knowledge proof generation and neural network inference, with C FFI bindings for integration with other languages/platforms.

Structure

• src/
  • core/ - Core logic for ZK proof, tensor operations, error handling, and callbacks.
  • ffi.rs - C FFI bindings for the Rust core.
  • lib.rs - Library entry point.
• headers/zk_neural_rust_core.h - C header generated for FFI.
• test.c - Example C usage/testing.
• scripts/ - Build, test, and packaging scripts.

Features

• Generate witnesses and proofs via user-provided callbacks.
• TensorFlow Lite model inference via TensorInvoker.
• C FFI for all major operations.
• Designed for cross-platform use, including iOS (see scripts).

Usage

Rust

Add as a dependency or use as a library crate.

C FFI

1. Build the Rust library as a shared or static library.
2. Use the generated zk_neural_rust_core.h header.
3. See test.c for example usage.

Scripts

• scripts/build_ios.sh - Build for iOS targets.
• scripts/create_xcframework.sh - Package as an XCFramework for Apple platforms.
• scripts/create_tflitec_xcframework.sh - Package TensorFlow Lite C dylib as an XCFramework.
• scripts/run_test.sh - Build and run the C test.
• scripts/update_header.sh - Update the C header file from Rust definitions.

Example

See test.c for a demonstration of witness and proof generation via FFI.

License

MIT License. See LICENCE for details.