 slaren
|
f30ea47a87
llama : add pipeline parallelism support (#6017)
|
hai 1 ano |
 Georgi Gerganov
|
5b09797321
ggml : remove old quantization functions (#5942)
|
hai 1 ano |
 0cc4m
|
61d1c88e15
Vulkan Improvements (#5835)
|
hai 1 ano |
 Michael Podvitskiy
|
9fa2627347
ggml : introduce ggml_status (ggml/750)
|
hai 1 ano |
 ddpasa
|
c2224f003b
ggml-vulkan: fix VULKAN_CHECK_RESULTS flag, which was previously broken (#5813)
|
hai 1 ano |
 Eve
|
317709b2a8
make portability_enumeration_ext apple only (#5757)
|
hai 1 ano |
 UEXTM.com
|
5f70671856
Introduce backend GUIDs (ggml/743)
|
hai 1 ano |
|
Georgi Gerganov
|
ab336a9d5e
code : normalize enum names (#5697)
|
hai 1 ano |
|
0cc4m
|
bb9dcd560a
Refactor validation and enumeration platform checks into functions to clean up ggml_vk_instance_init()
|
hai 1 ano |
|
0cc4m
|
f50db6ae0b
Add check for VK_KHR_portability_enumeration for MoltenVK support
|
hai 1 ano |
 Mathijs de Bruin
|
d8c054517d
Add preprocessor checks for Apple devices.
|
hai 1 ano |
|
Mathijs de Bruin
|
42f664a382
Resolve ErrorIncompatibleDriver with Vulkan on MacOS.
|
hai 1 ano |
|
Georgi Gerganov
|
5bf2b94dd4
cmake : fix VULKAN and ROCm builds (#5525)
|
hai 1 ano |
 Neuman Vong
|
704359e299
vulkan: Find optimal memory type but with fallback (#5381)
|
hai 1 ano |
 Sergio López
|
c88c74f967
vulkan: only use M-sized matmul on Apple GPUs (#5412)
|
hai 1 ano |
|
0cc4m
|
44fbe34360
Fix Vulkan crash on APUs with very little device memory (#5424)
|
hai 1 ano |
|
0cc4m
|
ee1628bdfe
Basic Vulkan Multi-GPU implementation (#5321)
|
hai 1 ano |
|
0cc4m
|
e920ed393d
Vulkan Intel Fixes, Optimizations and Debugging Flags (#5301)
|
hai 1 ano |
|
0cc4m
|
4d0924a890
Vulkan Phi Fix for AMD Proprietary Drivers (#5260)
|
hai 1 ano |
|
0cc4m
|
f8e9140cb4
Vulkan Fixes (#5223)
|
hai 1 ano |
|
0cc4m
|
2256f36b79
Vulkan Windows APU Memory Handling (#5199)
|
hai 1 ano |
|
0cc4m
|
2307523d32
ggml : add Vulkan backend (#2059)
|
%!s(int64=2) %!d(string=hai) anos |
