# ggerganov/llama.cpp b9045: b9045 — retained record

> Retained github record from github:ggerganov/llama.cpp.

Canonical: https://highsignal.app/data/records/29ad2c1c16f34961

[← sources](https://highsignal.app/data)


source: [github](https://highsignal.app/data/github) (github:ggerganov/llama.cpp)  published 2026-05-06 13:33 ingested 2026-05-07 08:19

[original source ↗](https://github.com/ggml-org/llama.cpp/releases/tag/b9045)

## extract

<details open>

mtmd: add granite-speech support (ibm-granite/granite-4.0-1b-speech) (#22101)

* mtmd: add granite-speech support (ibm-granite/granite-4.0-1b-speech)

Conformer encoder with Shaw relative position encoding,
QFormer projector, log-mel spectrogram with frame stacking.

Encoder uses GLU gating, folded batch norm, and SSM depthwise
conv. QFormer compresses encoder output via windowed
cross-attention (window=15, queries=3) into the LLM embedding
space.

Audio preprocessing: reflect-padded STFT, 80-bin mel filterbank,
dynamic range compression, 2x frame stacking (80->160 mel).

GGUF converter handles batch norm folding at export time,
fused K/V split, and Conv1d weight reshaping.

Tested against HF transformers reference: token-for-token match
on 30s/60s audio clips with greedy decoding.

* mtmd: rename gs_ prefixed tensors to generic/architecture names

* mtmd: use tensor_mapping.py for all granite_speech tensors

* convert: fold GraniteSpeechTextModel into GraniteModel

* mtmd: replace n_layer hack with explicit has_standard_layers flag

* mtmd: replace hardcoded magic numbers with GGUF hparams for granite speech

* mtmd: align KEY_A_ define spacing

* convert: register GraniteModel for GraniteSpeechForConditionalGeneration

* convert: fix ty type-check for GraniteSpeechMmprojModel registration

* mtmd: align TN_ define spacing

* mtmd: use generic layer loop for granite speech tensor loading

* mtmd: merge qformer_proj_layer into clip_layer

* mtmd: granite_speech remove redundant ggml_build_forward_expand on inputs

* mtmd: granite_speech add comment explaining why build_attn is not used

* mtmd: granite_speech hard-code eps in cpp, remove from GGUF metadata

* gguf: add spacing between granite_speech tensor mapping blocks

* mtmd: make generic audio layer_norm_eps read optional

* mtmd: granite_speech keep encoder eps in GGUF, only hard-code projector eps

* mtmd: align defines and struct fields in clip-impl.h and clip-model.h

* mtmd: fix alignment and ordering issues across granite speech files

* convert: granite_speech use filter_tensors instead of modify_tensors for skipping

</details>

**macOS/iOS:**
- [macOS Apple Silicon (arm64)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-macos-arm64.tar.gz)
- [macOS Apple Silicon (arm64, KleidiAI enabled)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-macos-arm64-kleidiai.tar.gz)
- [macOS Intel (x64)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-macos-x64.tar.gz)
- [iOS XCFramework](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-xcframework.zip)

**Linux:**
- [Ubuntu x64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-ubuntu-x64.tar.gz)
- [Ubuntu arm64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-ubuntu-arm64.tar.gz)
- [Ubuntu s390x (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-ubuntu-s390x.tar.gz)
- [Ubuntu x64 (Vulkan)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-ubuntu-vulkan-x64.tar.gz)
- [Ubuntu arm64 (Vulkan)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-ubuntu-vulkan-arm64.tar.gz)
- [Ubuntu x64 (ROCm 7.2)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-ubuntu-rocm-7.2-x64.tar.gz)
- [Ubuntu x64 (OpenVINO)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-ubuntu-openvino-2026.0-x64.tar.gz)
- [Ubuntu x64 (SYCL FP32)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-ubuntu-sycl-fp32-x64.tar.gz)
- [Ubuntu x64 (SYCL FP16)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-ubuntu-sycl-fp16-x64.tar.gz)

**Android:**
- [Android arm64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-android-arm64.tar.gz)

**Windows:**
- [Windows x64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-win-cpu-x64.zip)
- [Windows arm64 (CPU)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-win-cpu-arm64.zip)
- [Windows x64 (CUDA 12)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-win-cuda-12.4-x64.zip) - [CUDA 12.4 DLLs](https://github.com/ggml-org/llama.cpp/releases/download/b9045/cudart-llama-bin-win-cuda-12.4-x64.zip)
- [Windows x64 (CUDA 13)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-win-cuda-13.1-x64.zip) - [CUDA 13.1 DLLs](https://github.com/ggml-org/llama.cpp/releases/download/b9045/cudart-llama-bin-win-cuda-13.1-x64.zip)
- [Windows x64 (Vulkan)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-win-vulkan-x64.zip)
- [Windows x64 (SYCL)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-win-sycl-x64.zip)
- [Windows x64 (HIP)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-win-hip-radeon-x64.zip)

**openEuler:**
- [openEuler x86 (310p)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-310p-openEuler-x86.tar.gz)
- [openEuler x86 (910b, ACL Graph)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-910b-openEuler-x86-aclgraph.tar.gz)
- [openEuler aarch64 (310p)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-310p-openEuler-aarch64.tar.gz)
- [openEuler aarch64 (910b, ACL Graph)](https://github.com/ggml-org/llama.cpp/releases/download/b9045/llama-b9045-bin-910b-openEuler-aarch64-aclgraph.tar.gz)

## retained text

No document body was retained for this record — only its metadata was stored.
