ggerganov/llama.cpp

Pull Requests Commits

docs: Clarify the endpoint that webui uses (#17001)

openingnow committed 58 days ago

Verified fd2f84f4

model : add openPangu-Embedded (#16941)

Lpzhan931 committed 58 days ago

Verified 9f052478

ggml webgpu: minor set rows optimization (#16810)

reeselevine committed 58 days ago

Verified 03ea0417

ggerganov committed 58 days ago

cdabeb2c

ggml : fix conv2d_dw SVE path (ggml/1380)

ggerganov committed 58 days ago

852ce518

CUDA: update ops.md (#17005)

mnehete32 committed 58 days ago

Verified 9aa63374

opencl: update doc (#17011)

lhez committed 58 days ago

Verified 5e90233b

refactor: replace sprintf with snprintf for safer string handling in dump functions (#16913)

chraac committed 58 days ago

Verified a5c07dcd

vulkan: remove the need for the dryrun (#16826)

jeffbolznv committed 58 days ago

Verified ad51c0a7

server : do context shift only while generating (#17000)

ggerganov committed 58 days ago

Verified 66d8eccd

readme : update hot topics (#17002)

ggerganov committed 59 days ago

Verified afd35324

ggml-cpu : bicubic interpolation (#16891)

Acly committed 59 days ago

Verified cc98f8d3

ci : apply model label to models (#16994)

CISC committed 59 days ago

Verified d9458343

chore : fix models indent after refactor (#16992)

CISC committed 59 days ago

Verified b164259b

Fix garbled output with REPACK at high thread counts (#16956)

NoahOksuz committed 59 days ago

Verified 1f5accb8

CUDA: avoid mul + bias fusion when doing fusion (#16935)

am17an committed 59 days ago

Verified 2759ccdb

opencl: support imrope (#16914)

lhez committed 59 days ago

Verified c5023daf

fix: Viewing multiple PDF attachments (#16974)

allozaur committed 59 days ago

Verified e7da30b5

model-conversion : pass config to from_pretrained (#16963)

danbev committed 60 days ago

Verified ed8aa633

server : add props.model_alias (#16943)

ggerganov committed 60 days ago

Verified 48bd2650

ggml: CUDA: add head size 72 for flash-attn (#16962)

theo77186 committed 60 days ago

Verified 622cd010

mtmd: add --image-min/max-tokens (#16921)

ngxson committed 60 days ago

Verified 070ff4d5

mtmd: pad mask for qwen2.5vl (#16954)

ngxson committed 60 days ago

Verified bf7b0c97

ggml : LoongArch fixes (#16958)

MQ-mengqing committed 60 days ago

Verified fcfce040

sync: minja (glm 4.6 & minmax m2 templates) (#16949)

ochafik committed 60 days ago

Verified ee3a5a10

SYCL: optimized repeat_back kernel (3× fewer asm instructions, 2× faster)Feature/sycl repeat back opt (#16869)

shani-f committed 60 days ago

Verified 7e994168

feat(webui): improve LaTeX rendering with currency detection (#16508)

srogmann committed 60 days ago

Verified bcfa8762

test-backend-ops : fix segfault in moe-expert-reduce test in support mode and coverage (#16936)

sbera77 committed 60 days ago

Verified a2054e3a

ci : disable failing riscv cross build (#16952)

CISC committed 60 days ago

Verified dd528680

model: add Janus Pro for image understanding (#16906)

ravenouse committed 60 days ago

Verified 6b9a5242

Newer Older