On-Device, Local LLM iOS App Example

Source code to run Microsoft's Phi 3.5 Min 4K model completely on device. Essentially a mini ChatGPT but fully running on your phone/device. Requires knowledge of Xcode, and Apple Silicon hardware (iPhone, iPad, Vision Pro, M series Mac) with about 3+ GB of RAM. You'll run the app, the model will download, and then you can use the app. Includes current fixes to issues in open source software. Uses the MLX frameworks. Note that this requires an actual device, as opposed to a simulator.