Update README.md

Author: jllllll

README.md

@@ -2,15 +2,15 @@
 Wheels for [llama-cpp-python](https://github.com/abetlen/llama-cpp-python) compiled with cuBLAS support.
 
 Requirements:
-- Windows x64, Linux x64, or MacOS 11.7+
+- Windows x64, Linux x64, or MacOS 11.0+
 - CUDA 11.6 - 12.2
 - CPython 3.8 - 3.11
 
 llama.cpp, and llama-cpp-python by extension, has migrated to using the new GGUF format and has dropped support for GGML.
 This applies to version 0.1.79+.
 
 ROCm builds for AMD GPUs: https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/tag/rocm  
-Metal builds for MacOS 11.7+: https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/tag/metal
+Metal builds for MacOS 11.0+: https://github.com/jllllll/llama-cpp-python-cuBLAS-wheels/releases/tag/metal
 
 Installation instructions:
 ---