**What Happened:**
The publication of a new AI model named Command A+ (218B MoE) running on Apple Silicon has been announced. The model, originally developed by Cohere and based on the ML-5 architecture with 218 billion parameters, was ported to run on Apple silicon hardware using the MLX framework. This effort resulted in a pull request being opened for review at ml-explore/mlx-lm repository.
**Why It Matters:**
This development is significant because it demonstrates the capability of running large-scale language models (LLMs) like Command A+ on Apple Silicon, which typically use ARM processors rather than Intel x86. This portation not only showcases the versatility and adaptability of these LLMs but also highlights how they can be optimized for different hardware architectures. Moreover, it opens up possibilities for developers to run such models in environments that are more energy-efficient and have specific security features inherent in Apple silicon.
– Command A+ is now running on Apple Silicon.
– The model was ported using the MLX framework developed by Cohere.
– A pull request has been opened for review at ml-explore/mlx-lm repository, indicating further development and potential improvements are underway.
Originally published at reddit.com. Curated by AI Maestro.
Stay ahead of AI. Get the most important stories delivered to your inbox — no spam, no noise.
