Αυτή είναι παλαιότερη έκδοση. Η τελευταία έκδοση είναι η b9994. llama.cpp b9994 · Κυκλοφορία:
Αρχεία εγκατάστασης b9893
| Αρχιτεκτονική | Τύπος | Εμβέλεια | Μέγεθος αρχείου | Λήψη | SHA256 |
|---|---|---|---|---|---|
| 64-bit | ZIP → Φορητό | — | — | github.com | VirusTotal |
Ο σύνδεσμος και το SHA256 προέρχονται από το κοινοτικό manifest του Windows Package Manager. Συγκρίνετε το hash του ληφθέντος αρχείου πριν το εκτελέσετε.
Σημειώσεις έκδοσης
opencl: general flash attention decode performance optimizations (#25366)
- opencl: vec flash-attention decode kernels for f16/q8_0/q4_0 KV
- opencl: improve non FA KQ mv kernels
- opencl: tweaks for multiquery FA
- opencl: some tweaks for FA q1 kernels
- opencl: FA with DK=DV=512 for gemma-4
- opencl: various fixes
- opencl: cleanup
- opencl: fix FA decode crash for DK=512 (gemma-4)
The DK=512 decode-only program does not create the f32_f16 prefill
kernel, so the compiled check in ensure_fa_variant never hit and
supports_op gave inconsistent answers for the same op. block_n is also
unset for DK=512 decode; guard it to avoid an out-of-range read at
dispatch.
- opencl: run DK=512 FA decode on CPU
DK=512 decode is bandwidth-bound and faster on the CPU than the GPU,
increasingly so with depth. Decline it in supports_op; prefill stays on the GPU.
- opencl: compile MQ_GQA=8 FA kernels in a minimal program
The full program compiled with -D MQ_GQA=8 runs the Adreno compiler out
of memory at DK>=256. Only the vec_mq kernels are used from this
program, so compile it with FA_MQ_ONLY, which excludes everything else.
Also include the program name in the compile error log.
- opencl: remove stray token in flash_attn_f32_f16.cl
A stray "." broke the f32_f16 program build.
- opencl: split f16-KV FA decode finer (FD_KV_PER_SPLIT_F16)
The 2048 default under-fills the GPU on single-query f16-KV decode;
use 512 for f16 KV to get more splits. Quantized KV keeps 2048.
Co-authored-by: Li He lih@qti.qualcomm.com
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 7.2)
- 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 x64 (Vulkan)
- Windows x64 (OpenVINO)
- Windows x64 (SYCL)
- Windows x64 (HIP)
openEuler:
- DISABLED
- openEuler x86 (310p)
- openEuler x86 (910b, ACL Graph)
- openEuler aarch64 (310p)
- openEuler aarch64 (910b, ACL Graph)
UI:
- UI