ggml/llama.cpp releases: b10989

TL;DR

The ggml/llama.cpp project released version b10989, which includes an update to enable AllReduce for ROCm (#27825). This update is significant as it enhances the project's compatibility and performance on AMD GPUs.

Detailed Summary

The ggml/llama.cpp project released version b10989, which includes an update enabling AllReduce for ROCm (#27825). This update is part of the broader efforts to enhance the performance and compatibility of the llama.cpp library on AMD's ROCm platform. The release impacts users looking to leverage GPU acceleration with this library on ROCm-supported systems.

Key Points

• • HIP: Enable AllReduce for ROCm (#27825)
• • Website: <https://llama.app>
• • Attestations: <https://github.com/ggml-org/llama.cpp/attestations/47762387>