| .. |
|
ggml-alloc.h
|
b1f3a6e5db
llama: automatically set parameters not set by the user in such a way that maximizes GPU utilization (#16653)
|
1 mēnesi atpakaļ |
|
ggml-backend.h
|
b1f3a6e5db
llama: automatically set parameters not set by the user in such a way that maximizes GPU utilization (#16653)
|
1 mēnesi atpakaļ |
|
ggml-blas.h
|
ae8de6d50a
ggml : build backends as libraries (#10256)
|
1 gadu atpakaļ |
|
ggml-cann.h
|
ae8de6d50a
ggml : build backends as libraries (#10256)
|
1 gadu atpakaļ |
|
ggml-cpp.h
|
4254bb4951
ggml : fix ggml_gallocr_ptr type (ggml/1205)
|
8 mēneši atpakaļ |
|
ggml-cpu.h
|
51604435e8
ggml-cpu : fix RISC-V Q4_0 repack select and RVV feature reporting (#17951)
|
1 mēnesi atpakaļ |
|
ggml-cuda.h
|
ae8de6d50a
ggml : build backends as libraries (#10256)
|
1 gadu atpakaļ |
|
ggml-hexagon.h
|
63d2fc46e1
Add experimental ggml-hexagon backend for the Hexagon NPU (#16547)
|
2 mēneši atpakaļ |
|
ggml-metal.h
|
0320ac5264
metal : refactor + optimize v2 (#15995)
|
4 mēneši atpakaļ |
|
ggml-opencl.h
|
a76c56fa1a
Introducing experimental OpenCL backend with support for Qualcomm Adreno GPUs (#10693)
|
1 gadu atpakaļ |
|
ggml-opt.h
|
5cdb27e091
finetune: SGD optimizer, more CLI args (#13873)
|
5 mēneši atpakaļ |
|
ggml-rpc.h
|
8160b38a5f
rpc : fix alloc size logic (#17116)
|
1 mēnesi atpakaļ |
|
ggml-sycl.h
|
ae8de6d50a
ggml : build backends as libraries (#10256)
|
1 gadu atpakaļ |
|
ggml-vulkan.h
|
c2a67efe38
vulkan: Make Vulkan optional at runtime (#11493). (#11494)
|
11 mēneši atpakaļ |
|
ggml-webgpu.h
|
21c021745d
ggml: Add initial WebGPU backend (#14521)
|
6 mēneši atpakaļ |
|
ggml-zdnn.h
|
264f1b5187
zdnn: refactor codebase + add docs (#16178)
|
3 mēneši atpakaļ |
|
ggml-zendnn.h
|
017761daf5
ggml-zendnn : add ZenDNN backend for AMD CPUs (#17690)
|
1 mēnesi atpakaļ |
|
ggml.h
|
45ada635f0
Corrected Implementation of Qwen3Next Support
|
3 nedēļas atpakaļ |
|
gguf.h
|
53ff6b9b9f
GGUF: C++ refactor, backend support, misc fixes (#11030)
|
1 gadu atpakaļ |
Çetin
45ada635f0
Corrected Implementation of Qwen3Next Support