🎬+🎶 llama.cpp [834a243]

🛠️ Latest changes: • ggml-webgpu: Enable NVIDIA self-hosted CI (PR-22976) • ggml-webgpu: makes the flash attn vec path subgroup-aware (PR-23040) • fix: Propagate version tag to WebUI asset download in self-hosted CI (PR-23051) • contributing: new contributors should not submit trivial fixes (PR-23045) • webui: Move static build output from repo code to HF Bucket (PR-22937) • logs : reduce (PR-23021) • ggml-cpu: Add IME2 Instruction Support for the SpacemiT Backend (PR-22863) • docker : revert stable version of intel compute-runtime (PR-22968) • unicode,test: add Qwen3.5 non-backtracking tokenizer handler and regr… (PR-22110) • vulkan: fix matmul integer pipeline selection (PR-23005) 📊 Stats: 876 files changed, 97591 insertions(+), 57166 deletions(-) 🔢 Total commits: 286 💻 Codebase: 669,583 lines (C++ 52.9%, C/C++ Header 24.1%, Python 8.4%) 🔗 Repository: https://github.com/ggml-org/llama.cpp.git 🎶 Soundtrack: Generated chiptune ________________________________________________ 🤖 Generated by: chaos 🖥️ Hardware: NVIDIA T400, 570.172.08 ⏱️ Render time: 45:55 #llama #cpp #GitVisualization #CodeHistory