ggerganov/llama.cpp

Pull Requests Commits

llama : avoid double tensor copy when saving session to buffer

compilade committed 1 year ago

9329953a

llama : avoid useless copies in dummy session writer

compilade committed 1 year ago

dca7ad86

make : use C compiler to build metal embed object (#8899)

slaren committed 1 year ago

Verified 15fa07a5

ggml-backend : fix async copy from CPU (#8897)

slaren committed 1 year ago

Verified be55695e

[SYCL] Updated SYCL device filtering (#8901)

OuadiElfarouki committed 1 year ago

Verified 0478174d

CUDA/HIP: fix tests/test-backend-ops (#8896)

JohannesGaessler committed 1 year ago

Verified a8dbc6f7

llama-bench : add support for getting cpu info on Windows (#8824)

kylo5aby committed 1 year ago

Verified 506122d8

quantize : update usage comment in quantize.cpp (#8889)

danbev committed 1 year ago

Verified 725e3d94

typo correction (#8891)

Nexesenex committed 1 year ago

Verified 31958546

server : add lora hotswap endpoint (WIP) (#8857)

ngxson committed 1 year ago

Verified 1e6f6554

CUDA: fix padding logic for FP16/FP32 (#8884)

JohannesGaessler committed 1 year ago

Verified 641f5dd2

simple : update name of executable to llama-simple (#8885)

danbev committed 1 year ago

Verified 5f4dcb1e

cmake : Link vulkan-shaders-gen with pthreads (#8835)

Patater committed 1 year ago

Verified db20f50c

[Vulkan] Fix compilation of `vulkan-shaders-gen` on w64devkit after `e31a4f6` (#8880)

MaggotHATE committed 1 year ago

Verified efda90c9

contributing : add note about write access

ggerganov committed 1 year ago

Verified 0bf16de0

ggml : add epsilon as a parameter for group_norm (#8818)

MollySophia committed 1 year ago

Verified 2d5dd7bb

convert : add support for XLMRoberta embedding models (#8658)

iamlemec committed 1 year ago

Verified cdd1889d

[CANN]: Fix ggml_backend_cann_buffer_get_tensor (#8871)

MengqingCao committed 1 year ago

Verified c21a8964

[SYCL] correct cmd name (#8877)

arthw committed 1 year ago

Verified d4ff8471

common : Changed tuple to struct (TODO fix) (#8823)

Septa2112 committed 1 year ago

Verified 0a4ce786

cann: fix buffer_num and runtime speed slowly error (#8865)

wangshuai09 committed 1 year ago

Verified bc0f887e

readme : add ramalama to the availables UI (#8811)

ericcurtin committed 1 year ago

Verified b42978e7

ggml : fix overflows in elu function (#8866)

jart committed 1 year ago

Verified b9dfc25c

py: Add more authorship metadata from model card (#8810)

mofosyne committed 1 year ago

Verified 1ef14b30

Stop the generation when <|eom_id|> token is encountered - needed for Llama 3.1 tool call support (#8858)

fairydreaming committed 1 year ago

Verified d3f0c716

cmake: fix paths for vulkan shaders compilation on Windows (#8573)

stduhpf committed 1 year ago

Verified e31a4f67

readme : update model list (#8851)

BarfingLemurs committed 1 year ago

Verified 400ae6f6

llama : better replace_all (#8852)

ggerganov committed 1 year ago

Verified f1ea5146

vulkan : fix Qantized Mat-Vec Mul on AMD GPUs for ncols < 64 (#8855)

0cc4m committed 1 year ago

Verified 064cdc26

ggerganov committed 1 year ago

5587e57a

Older