llama.cpp Releases
609 releases of ggml-org/llama.cpp
- b9891
A new operation `col2im_1d` has been added to the Metal backend, supporting floating-point formats (f32, f16, bf16), mirroring functionality available in CPU and CUDA backends.
Jul 6, 2026
- b9890
A significant refactoring of the CUDA backend removes the `-sm row` option and reworks the cuBLAS implementation. This includes fixes for AMD CDNA compatibility with BF16 and corrections to memory stride handling.
Jul 6, 2026
- b9888
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.
Jul 6, 2026
- b9886
The CPU backend has been optimized for ARM architecture by utilizing the UE4M3 Look-Up Table (LUT) for ARM NVFP4 dot product operations, aiming to improve computational efficiency.
Jul 6, 2026
- b9885
Tiled matrix multiplication has been enabled for AIX systems by reducing local stack buffer footprints to avoid segmentation faults. This change delivers approximately 2x performance improvements for specific model types on the platform.
Jul 6, 2026
- b9884
A critical bug fix for the Vulkan backend addresses a 32-bit integer overflow occurring in the CEIL_DIV macro, preventing potential crashes or incorrect calculations on 32-bit platforms.
Jul 6, 2026
- b9882
The build scripts have been updated to utilize the Hugging Face (HF) token when downloading UI assets, ensuring authenticated access to private or restricted resources.
Jul 6, 2026
- b9881
This release enables the `-ffast-math` compiler flag for HIP (AMD GPU) builds, which optimizes floating-point arithmetic for improved performance on AMD hardware.
Jul 6, 2026
- b9878
This release fixes stale tensor-split parameters for draft models, specifically correcting metadata for GQA attention. It also includes a revert of a recent code tidying commit and a backend meta asset hack.
Jul 5, 2026
- b9877
This release introduces a safety check that aborts the program execution if a multi-buffer configuration is detected, likely to prevent undefined behavior or crashes in specific backend scenarios.
Jul 5, 2026
- b9876
Fixes a crash when using tensor parallelism (-sm tensor) combined with CPU-offloaded MoE experts (-ncmoe). The fix moves the split-state lookup before the contiguity assert.
Jul 5, 2026
- b9874
Implements a CUDA version of the concat operation for quantized tensors. This mirrors the CPU fix from b9871 and improves GPU backend performance.
Jul 5, 2026
- b9873
Adds a guard for K/V rotation inputs when buffers are unallocated. This prevents a crash (GGML_ASSERT) in DFlash speculative decoding when buffers are NULL.
Jul 4, 2026
- b9871
Fixes a broken CPU implementation for concatenating quantized tensors. The previous implementation likely failed or produced incorrect results for quantized types.
Jul 4, 2026
- b9870
Fixes long reasoning loops in chat by trimming messages sent to the StepFun parser. The trimming is applied before rendering to content parts.
Jul 3, 2026
- b9867
Extends support for speculative decoding in DFlash by adding the spec-draft-p-min parameter. Includes guards for n_min to ensure valid configurations.
Jul 3, 2026
- b9866
Enables topk-moe fusion for 288 experts in CUDA, previously only supported for power-of-2 counts or 576. This improves decode performance for large MoE models.
Jul 3, 2026
- b9864
Enhances server and UI stability for SSE streams by adding silent ping mechanisms. It introduces a configurable ping interval to prevent disconnects during slow prefill operations.
Jul 3, 2026
- b9862
This release optimizes CUDA kernels by removing redundant copies after gated_delta_net operations. It writes recurrent state snapshots directly into the recurrent cache, reducing kernel calls for MTP draft lengths.
Jul 3, 2026
- b9861
Updates the vendor library cpp-httplib to version 0.49.0. This is a dependency update with no functional changes to the core inference engine.
Jul 3, 2026