ngxson/llama.cpp

Pull Requests Commits

ngxson committed 1 year ago

f01bc37c

ngxson committed 1 year ago

98da78b2

ngxson committed 1 year ago

e91c5780

clean up struct def

ngxson committed 1 year ago

c58a332f

Merge branch 'master' into xsn/lora_server_hotswap

ngxson committed 1 year ago

21cb1338

ngxson committed 1 year ago

aa3cea0f

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

vulkan : implement Stable Diffusion operators (ggml/904)

0cc4m committed 1 year ago

a3738b2f

ggml : move c parameter comment to ggml_rope_ext (ggml/901)

danbev committed 1 year ago

655858ac

cann: support q4_0 model (#8822)

wangshuai09 committed 1 year ago

Verified c02b0a8a

Install curl in runtime layer (#8693)

bsquizz committed 1 year ago

Verified 0d6fb52b

Server: Don't ignore llama.cpp params (#8754)

ardfork committed 1 year ago

Verified 978ba3d8

ngxson committed 1 year ago

e7d9abf3

handle lora_no_apply

ngxson committed 1 year ago

42960ec7

server : add lora hotswap endpoint

ngxson committed 1 year ago

e5c2d167

batched-bench : handle empty `-npl` (#8839)

cunnie committed 1 year ago

Verified ecf6b7f2

baby-llama : remove duplicate vector include

danbev committed 1 year ago

01aae2b4

flake.lock: Update (#8847)

ggerganov committed 1 year ago

Verified 4b77ea95

ggml : reading the runtime sve config of the cpu (#8709)

jdomke committed 1 year ago

Verified 76614f35

Fix conversion of unnormalized BF16->BF16 weights (#7843)

CISC committed 1 year ago

Verified b72c20b8

Older