 Diego Devesa
|
c5b0f4b5d9
llama : refactor model loader with backend registry (#10026)
|
1 year ago |
 Changyeon Kim
|
8f275a7c45
ggml: Add POOL2D OP for GPU acceleration to the Vulkan backend in the MobileVLM model. (#9763)
|
1 year ago |
|
Diego Devesa
|
f010b77a37
vulkan : add backend registry / device interfaces (#9721)
|
1 year ago |
|
Diego Devesa
|
dca1d4b58a
ggml : fix BLAS with unsupported types (#9775)
|
1 year ago |
 SRHMorris
|
b0915d5b51
vulkan : retry allocation with fallback flags (whisper/2451)
|
1 year ago |
|
Diego Devesa
|
c83ad6d01e
ggml-backend : add device and backend reg interfaces (#9707)
|
1 year ago |
 Radoslav Gerganov
|
00b7317e63
vulkan : do not use tensor->extra (#9407)
|
1 year ago |
 Salvatore Mesoraca
|
cb00020504
vulkan : mul_mat: fix UB with small warps (ggml/952)
|
1 year ago |
 Jeff Bolz
|
641002fba8
vulkan : multithread pipeline creation (ggml/963)
|
1 year ago |
|
Jeff Bolz
|
0de8b203f1
vulkan : fix build for GGML_VULKAN_RUN_TESTS, add TFLOPS to log (ggml/961)
|
1 year ago |
 Markus Tavenrath
|
89f9944981
Enable use to the rebar feature to upload buffers to the device. (#9251)
|
1 year ago |
 Johannes Gäßler
|
424c5d00a9
ggml/examples: add backend support for numerical optimization (ggml/949)
|
1 year ago |
 Georgi Gerganov
|
d6a04f872d
ggml : hide ggml_object, ggml_cgraph, ggml_hash_set (#9408)
|
1 year ago |
|
Markus Tavenrath
|
daa9623ab0
Overlap cmdbuffer creation and cmdbuffer execution in Vulkan backend by submitting smaller cmdbuffers early. (#9118)
|
1 year ago |
|
Salvatore Mesoraca
|
406c1a32a1
vulkan: add dryrun support to sin and cos ops (ggml/947)
|
1 year ago |
|
Salvatore Mesoraca
|
9cb9260861
vulkan: correctly report support for OP_CONT (ggml/946)
|
1 year ago |
|
Changyeon Kim
|
409dc4f8bb
ggml : fix build break for the vulkan-debug (#9265)
|
1 year ago |
|
Georgi Gerganov
|
231cff5f6f
sync : ggml
|
1 year ago |
|
Changyeon Kim
|
2f3c1466ff
llava: Add ACC OP for GPU acceleration to the Vulkan backend in the LLAVA CLIP model. (#8984)
|
1 year ago |
 0cc4m
|
5fd89a70ea
Vulkan Optimizations and Fixes (#8959)
|
1 year ago |
 Daniel Bevenius
|
06943a69f6
ggml : move rope type enum to ggml.h (#8949)
|
1 year ago |
|
Markus Tavenrath
|
7c5bfd57f8
Optimize Vulkan backend for better CPU performance and less GPU synchronization overhead. (#8943)
|
1 year ago |
 Matt Stephenson
|
70c0ea3560
whisper : use vulkan as gpu backend when available (whisper/2302)
|
1 year ago |
|
0cc4m
|
a3738b2fa7
vulkan : implement Stable Diffusion operators (ggml/904)
|
1 year ago |
 Tony Wasserka
|
203b7f1531
vulkan : initialize vk_buffer_struct members to VK_NULL_HANDLE (ggml/893)
|
1 year ago |
|
slaren
|
2b1f616b20
ggml : reduce hash table reset cost (#8698)
|
1 year ago |
|
0cc4m
|
751fcfc6c3
Vulkan IQ4_NL Support (#8613)
|
1 year ago |
|
0cc4m
|
bda62d7999
Vulkan MMQ Fix (#8479)
|
1 year ago |
|
Georgi Gerganov
|
f3f65429c4
llama : reorganize source code + improve CMake (#8006)
|
1 year ago |
