b9888
ggml-org/llama.cppb9888Jul 6, 2026by github-actions[bot]same day as b9886
What changed since b9886
CUDA Flash Attention now validates V-matrix types in addition to K-types, tightening type checks.
Improvements
- CUDA Flash Attention extends K-type validation to also cover V-types, improving type safety.
- Stricter type compatibility checks in the CUDA Flash Attention path.
Upgrade notes
This is a small incremental build over b9886 with no config or API changes noted. CUDA users relying on Flash Attention benefit from stricter type validation; no migration steps are required.
Analyzed Oct 28, 2026
Upgrading?
Release notes summary
The CUDA backend for Flash Attention has been updated to validate the types of both K (Key) and V (Value) matrices, ensuring stricter type compatibility checks.
Key Highlights
- Extend K-type validation to V-types in Flash Attention
- Improve CUDA Flash Attention type safety
Release Notes (excerpt)
<details open> CUDA: extend K-type validation to V-types for flash attention (#24403) * CUDA: extend K-type validation to V-types for flash attention * reorder </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)…
Published by ggml-org on GitHub.
Read the full release notesRelated Repositories
:electron: Build cross-platform desktop apps with JavaScript, HTML, and CSS
Godot Engine – Multi-platform 2D and 3D game engine
ClickHouse® is a real-time analytics database management system
Self-hosted game stream host for Moonlight.
DuckDB is a high-performance analytical database system