Change8

b8323

📦 llama-cppView on GitHub →

Summary

This release disables graph reuse when using pipeline parallelism for llama models and provides numerous pre-compiled binaries for macOS, Linux, Windows, and openEuler.