@@ -9,6 +9,7 @@ Inference of [LLaMA](https://arxiv.org/abs/2302.13971) model in pure C/C++
**Hot topics:**
+- Roadmap June 2023: https://github.com/ggerganov/llama.cpp/discussions/1729
- GPU support with Metal (Apple Silicon): https://github.com/ggerganov/llama.cpp/pull/1642
- High-quality 2,3,4,5,6-bit quantization: https://github.com/ggerganov/llama.cpp/pull/1684
- Multi-GPU support: https://github.com/ggerganov/llama.cpp/pull/1607