Ollama v0.30.0 Redesigns Architecture for direct llama.cpp and GGUF Support

Core summary

One sentence to understand this update

Ollama v0.30.0 introduces a significant architectural change to directly support llama.cpp, GGUF file format compatibility, and MLX for accelerated inference on Apple hardware.

Impact & opportunity

What this could mean

This update simplifies integration with `llama.cpp` models and GGUF, offering better performance for local LLMs, especially on Apple devices, enabling broader application development.