Starstruk Logo MarkStarstruk

b10338

ggml-org/llama.cppb10338Aug 10, 2026by github-actions[bot]

AI Summary

Critical bug fix for the model saver that was incorrectly saving MoE (Mixture of Experts) model parameters, causing models like Qwen2MoE to fail loading after a save/load cycle.

Key Highlights

  • Fixed expert shared/chunk FFN length key clobber in model-saver.
  • Resolves loading failures for shared-expert MoE models.
  • Added tests to verify roundtrip save/load for MoE architectures.

Full Release Notes

<details open>

model-saver : fix expert shared/chunk FFN length key clobber (#26693)

The saver called add_kv with LLM_KV_EXPERT_SHARED_FEED_FORWARD_LENGTH twice, the
second time passing n_ff_chexp. gguf_set_val_u32 removes-then-appends, so the second
call clobbers the first: the saved shared_feed_forward_length ends up as n_ff_chexp
(0 for every arch except GroveMoE), and expert_chunk_feed_forward_length is never
written at all.

So a save->load roundtrip of any MoE model with a shared expert loses n_ff_shexp. On
reload the arch falls back to n_ff for the shexp tensor shape, that no longer matches
the saved tensor, and the model FAILS to load. Hits qwen2moe, qwen3-next, granite-moe,
hunyuan-moe, ernie4.5, bailingmoe2, nemotron-h, and the other shared-expert MoEs.

Fix: the second call writes LLM_KV_EXPERT_CHUNK_FEED_FORWARD_LENGTH.

test-llama-archs: set expert_shared_feed_forward_length to a value distinct from n_ff
in the MoE setup so the roundtrip exercises it. Without the fix the reload fails on a
shexp tensor-shape mismatch; with it, every arch roundtrips clean.

</details>

**Website:**
- <https://llama.app>

**macOS/iOS:**
- [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-macos-arm64.tar.gz)
- macOS Apple Silicon (arm64, KleidiAI enabled) [DISABLED](https://github.com/ggml-org/llama.cpp/pull/23780)
- [macOS Intel (x64)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-macos-x64.tar.gz)
- [iOS XCFramework](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-xcframework.zip)

**Linux:**
- [Ubuntu x64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-ubuntu-x64.tar.gz)
- [Ubuntu arm64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-ubuntu-arm64.tar.gz)
- [Ubuntu s390x (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-ubuntu-s390x.tar.gz)
- [Ubuntu x64 (Vulkan)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-ubuntu-vulkan-x64.tar.gz)
- [Ubuntu arm64 (Vulkan)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-ubuntu-vulkan-arm64.tar.gz)
- [Ubuntu x64 (ROCm 7.2)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-ubuntu-rocm-7.2-x64.tar.gz)
- [Ubuntu x64 (OpenVINO)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-ubuntu-openvino-2026.2.1-x64.tar.gz)
- [Ubuntu x64 (SYCL FP32)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-ubuntu-sycl-fp32-x64.tar.gz)
- [Ubuntu x64 (SYCL FP16)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-ubuntu-sycl-fp16-x64.tar.gz)

**Android:**
- [Android arm64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-android-arm64.tar.gz)

**Windows:**
- [Windows x64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-win-cpu-x64.zip)
- [Windows arm64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-win-cpu-arm64.zip)
- [Windows arm64 (OpenCL Adreno)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-win-opencl-adreno-arm64.zip)
- [Windows x64 (CUDA 12)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-win-cuda-12.4-x64.zip) - [CUDA 12.4 DLLs](https://github.com/ggml-org/llama.cpp/releases/download/b10338/cudart-llama-bin-win-cuda-12.4-x64.zip)
- [Windows x64 (CUDA 13)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-win-cuda-13.3-x64.zip) - [CUDA 13.3 DLLs](https://github.com/ggml-org/llama.cpp/releases/download/b10338/cudart-llama-bin-win-cuda-13.3-x64.zip)
- [Windows x64 (Vulkan)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-win-vulkan-x64.zip)
- [Windows x64 (OpenVINO)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-win-openvino-2026.2.1-x64.zip)
- [Windows x64 (SYCL)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-win-sycl-x64.zip)
- [Windows x64 (HIP)](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-bin-win-hip-radeon-x64.zip)

**openEuler:**
- [DISABLED](https://github.com/ggml-org/llama.cpp/pull/23705)
- openEuler x86 (310p)
- openEuler x86 (910b, ACL Graph)
- openEuler aarch64 (310p)
- openEuler aarch64 (910b, ACL Graph)

**UI:**
- [UI](https://github.com/ggml-org/llama.cpp/releases/download/b10338/llama-b10338-ui.tar.gz)