b9912

ggml-org/llama.cppb9912Jul 8, 2026by github-actions[bot]same day as b9911

What changed since b9911

low risk

b9912 is a housekeeping release that only cleans up naming and spacing in the speculative decoding module, with no functional changes.

Improvements

  • Speculative decoding source now follows consistent naming and spacing conventions.
  • Pure code-quality cleanup with no behavioral or performance impact.

Upgrade notes

This is a cosmetic-only release; no config, API, or model changes are involved. Upgrading from b9911 should be a drop-in replacement, and the NVFP4 CUDA fusion work from b9911 remains intact. Rebuild only if you track source rather than binaries.

Analyzed Oct 30, 2026

Upgrading?

Release notes summary

A minor code cleanup release fixing naming conventions and spacing issues in the speculative decoding module.

Key Highlights

  • Fix naming and spacing in speculative decoding module
  • Code quality improvements

New Features

  • Code cleanup

Release Notes (excerpt)

<details open> spec : fix naming, spacing (#25410) </details> **macOS/iOS:** - macOS Apple Silicon (arm64) - macOS Apple Silicon (arm64, KleidiAI enabled) DISABLED - macOS Intel (x64) - iOS XCFramework **Linux:** - Ubuntu x64 (CPU) - Ubuntu arm64 (CPU) - Ubuntu s390x (CPU) - Ubuntu x64 (Vulkan) - Ubuntu arm64 (Vulkan) - Ubuntu x64 (ROCm 7.2) - Ubuntu x64 (OpenVINO) - Ubuntu x64 (SYCL FP32) - Ubuntu x64 (SYCL FP16) **Android:** - Android arm64 (CPU) **Windows:** - Windows x64 (CPU)…

Published by ggml-org on GitHub.

Read the full release notes

Related Repositories