ANEMLL (pronounced like "animal") is an open-source project focused on accelerating the porting of Large Language Models (LLMs) to tensor processors, starting with the Apple Neural Engine. It is designed to process Model Weights directly from Hugging Face models and convert them to the CoreML format for Apple Neural Engine.

I’m still unclear why so much AI software on Mac uses Metal rather than ANE. Is it just easier to port to Metal? Is the performance higher?

Apple's article from the past has more info https://machinelearning.apple.com/research/neural-engine-tra...

ANE is different architecture and it's not easy to port existing models to it. TPU vs GPU difference.