chore: bump llama.cpp to b10375

[leehack]

llama.cpp update

• Previous pin: b10333
• New pin: b10375
• Upstream release: https://github.com/ggml-org/llama.cpp/releases/tag/b10375
• Compare: ggml-org/llama.cpp@b10333...b10375

Upstream changelog

Release notes for b10375

Details

chat : tighten bare function parsing for Qwen models (#26793)

Website:

• https://llama.app

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 7.14)
• Ubuntu x64 (OpenVINO)
• Ubuntu x64 (SYCL FP32)
• Ubuntu x64 (SYCL FP16)

Android:

• Android arm64 (CPU)

Windows:

• Windows x64 (CPU)
• Windows arm64 (CPU)
• Windows arm64 (OpenCL Adreno)
• Windows x64 (CUDA 12) - CUDA 12.4 DLLs
• Windows x64 (CUDA 13) - CUDA 13.3 DLLs
• Windows arm64 (CUDA 13) (preview) - CUDA 13.4 DLLs
• Windows x64 (Vulkan)
• Windows x64 (OpenVINO)
• Windows x64 (SYCL)
• Windows x64 (ROCm 7.14)

openEuler:

• DISABLED
• openEuler x86 (310p)
• openEuler x86 (910b, ACL Graph)
• openEuler aarch64 (310p)
• openEuler aarch64 (910b, ACL Graph)

UI:

• UI

Commit range

Commits from b10333 to b10375 (first 80)

• ci: add pr-draft-label (#26801) (9369185)
• ui: degrade the working directory picker when file search is off (#26811) (74ce157)
• ggml-webgpu : refactor several wgsl files and simplify flash_attn wgsl. (#26134) (f401bb1)
• ci: fix the ctest sanitize runs (#26593) (aea252f)
• model-saver : fix expert shared/chunk FFN length key clobber (#26693) (0377426)
• server: gate the docker tools runtime tests on a real container run (#26826) (1e396e7)
• ui: Linting & Formatting scripts (#26819) (92d1bb0)
• readme : remove dev branches (#26832) (6ad4ab0)
• model : Granite-Switch Architecture (#25107) (157b81f)
• vendor : update cpp-httplib to 0.53.0 (#26821) (e23e944)
• model: add MTP support for Nemotron model (#26725) (7a20b41)
• ci: Add support for CUDA 13.4 ARM64 builds for Windows (#26650) (2e2d99c)
• llama: Restore quantization of mmprojs (#26818) (86c298f)
• vendor: sync subprocess.h and drop local patches (#26808) (4c6766f)
• chat : Align Laguna-S-2.1 chat template to huggingface (#26232) (a52077c)
• model: Muse Glimmer Support (#26841) (62bf73d)
• server: add more tool isolation support (ssh remote + podman rootless) (#26774) (4ae84de)
• ci : don't specify python version in server-sanitize for broader runner compatibility (#26840) (e5275f6)
• ui: UI/chat form follow ups (#26743) (4dee52f)
• ggml : require contiguous src for ROLL on CUDA and Metal (#25928) (f8def7f)
• ggml-cpu : fix CPU affinity mask being ignored on Android (#26838) (d2f8305)
• llama : support multi-output backend sampling (#25532) (dd1ea52)
• ci : target ROCm 7.14 for build and release (#25775) (0666ad2)
• opencl: transpose the K tile in local memory for FA prefill kernels (#26428) (689e227)
• Address review comment of PR 25532 (#26852) (030ebb5)
• ggml-webgpu: fix CI errors from #25025 and #25262 (#26566) (84f7129)
• common/peg : suppress incomplete escape sequences (#26780) (48d22e2)
• model : fix SWA not being enabled for EXAONE 4.5 (#26848) (14e78dd)
• tests : disable backend sampler hip multi output (#26878) (4801e3c)
• tests : clean-up server test, use tests.sh in ci (#26886) (b3df572)
• llama: add default load-mode auto, which avoids mmap on iGPUs (#26081) (153d324)
• tests : fix running server tests on windows (#26889) (9afff1b)
• model-conversion : use save_output_data for causual embeddings [no ci] (#26890) (1138b85)
• ci: hip-quality-check: update vgpr spill ignore list (#26859) (7044859)
• ui: fix context gauge for single-model usage (#25738) (8d274dd)
• mtmd: support pocket-tts (#26871) (6e62ba5)
• Dflash support for nemotron-3.5 (#26905) (cc078b4)
• convert : keep quantization scales for nemotron --mtp export (#26903) (5d16e81)
• requirements: use stable torch packages on s390x (#26864) (2468576)
• imatrix.cpp: Move finite check and only check touched experts (#26861) (38406d5)
• ci : add windows-rocm to check-release (#26897) (70dfba5)
• chat : tighten bare function parsing for Qwen models (#26793) (ba360ef)

Web bridge review focus

Please pay extra attention to upstream changes touching:

• WebGPU, WASM, Emscripten, pthreads, or memory64 build behavior
• ggml backend APIs used by the bridge
• model loading, tokenizer, chat template, context/state persistence, or cache semantics
• CMake/build flags that can affect the generated JS/WASM artifacts

Validation

• Emscripten build passed
• Browser WebGPU/state-persistence smoke passed
• Generated bridge artifacts include wasm32 and memory64 outputs
• No stale hard-coded llama.cpp tag remains in CI/publish defaults

Automation behavior

This PR is managed from the stable branch automation/bump-llama-cpp. If another llama.cpp release appears before merge, the scheduled workflow updates this same PR instead of opening a duplicate. The workflow skips if a non-automation PR already changes llama_cpp.version.

[Copilot]

Pull request overview

Updates the pinned upstream llama.cpp release tag consumed by this repository’s CI/build workflows (via llama_cpp.version), advancing the bridge to build against the newer upstream release.

Changes:

• Bump llama_cpp.version from b10333 to b10356.

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.