llama.cpp b11747 adds ROCm 10 and CUDA 13 support

How it developed

1. October 20, 2026llama.cpp ReleasesWhere it started

   llama.cpp b11747 adds ROCm 10 and CUDA 13 support

2. October 28, 2026llama.cpp Releases

   llama.cpp adds LiquidAI d1 models and CUDA 13 support

   llama.cpp adds native support for LiquidAI d1-2B and d1-omni-2B models

3. October 28, 2026llama.cpp Releases

   llama.cpp adds LiquidAI LFM3-Audio and Omni support

   llama.cpp adds native conversion for LiquidAI LFM3-Audio and Omni architectures with specific encoder convolution handling