ggerganov/llama.cpp

Pull Requests Commits

ui: wrap long markdown tokens instead of scrolling every block

ServeurpersoCom committed 3 days ago

e00f7120

test : pin the server mode in the add menu a11y story

allozaur committed 3 days ago

a9316a03

ui : render get info as a collapsible block with a table

allozaur committed 3 days ago

b5ef6884

ui : keep long tool call titles inside the message row

allozaur committed 3 days ago

2ef5ffad

ui : fit wide image attachments to the message width

allozaur committed 3 days ago

cca015c2

ui : use exact transition properties for markdown images

allozaur committed 3 days ago

d2f0ec4c

ui : scroll long inline content inside markdown blocks

allozaur committed 3 days ago

2db23c20

ui : wrap markdown tables in a scroll container

allozaur committed 3 days ago

35fc09b2

ui : let the chat column shrink below its content width

allozaur committed 3 days ago

1600819c

opencl: add bin kernel `kernel_gemm_noshuffle_q6_k_f32_32b_trans_ila_a8_bin` (#28678)

shaofeiqi committed 3 days ago

Verified ec928150

ggml-cpu: add F16 input to the FWHT (#27779)

bri-prism committed 4 days ago

Verified 4fea119d

ggml : check for allocation failures to prevent crashes (#28149)

sanmai committed 4 days ago

Verified 5b335f41

Model-Saver: Write the SWA pattern, 15 more architectures roundtrip (#29042)

ServeurpersoCom committed 4 days ago

Verified 542348a3

ci: change ubuntu-latest to ubuntu-24.04 (#29079)

taronaeo committed 4 days ago

Verified d663dd3f

ggml-webgpu: fix supports_op condition for GET_ROWS (#28978)

yomaytk committed 4 days ago

Verified 44be98f0

ggml : handle graph buffer reservation failure (#26070)

FaiChou committed 4 days ago

Verified 911f6cdc

vulkan: add IQ3_S MMQ matmul kernels (#28822)

vargad committed 4 days ago

Verified bbd488c4

vocab : add ufakzeka pre-tokenizer (#29033)

stfurkan committed 4 days ago

Verified dc85f89c

cmake : fix build when GGML_CPU=OFF and GGML_CUDA=ON (#29026)

iamfobey committed 4 days ago

Verified 8ed1a55e

gguf-py: fix Q8_1 block size in GGML_QUANT_SIZES (2+2+32) (#29036)

YangHong7 committed 4 days ago

Verified bb11ebb6

ci : disable GHA cache for copilot (#29068)

CISC committed 4 days ago

Verified f03cf3e9

ci : bump android-actions/setup-android to 4.0.4 (#29065)

CISC committed 4 days ago

Verified bdcbaaf6

vulkan: raise the hoisted row-id limit for mul_mat_id from 256 to 512 experts (#28501)

drluoto committed 4 days ago

Verified 5c53396b

ci : add missing evict-old-files (#29041)

CISC committed 4 days ago

Verified 972d2313

rpc : skip ACCEL devices (#29020)

pcuenca committed 4 days ago

Verified c77ae695

model : skip gate_up_exps if TENSOR_SKIP is set (#29014)

dfriehs committed 5 days ago

Verified b49650ad

model : extend Nemotron MTP support (#29018)

frozenblade1224 committed 5 days ago

Verified 70761804

openvino : Update OpenVINO to 2026.4;fix clangd,MSVC warnings; (#29009)

ravi9 committed 5 days ago

Verified ebbb1852

ui: fix removed reasoning menu in single model mode on desktop (#27985)

benbenbenbenbenben committed 5 days ago

Verified 4ff829ec

vulkan: split buffers and debug code into separate files, add shared headers (#28732)

0cc4m committed 5 days ago

Verified f172be75

Older