ggml-org/llama.cpp

Pull Requests Commits

ngxson committed 1 hour ago

329dbff2

fix preproc (both k2.5 and k3.1) and projector mlp

ngxson committed 1 hour ago

637bacb3

mtmd: support kimi k3.1 vision

ngxson committed 1 hour ago

36985f11

mtmd: support gemma 4 input audio stream (#31965)

ngxson committed 1 hour ago

Verified 20c2e966

mtmd: add mtmd_helper_resampler (#31964)

ngxson committed 2 hours ago

Verified 3d94ffe4

ui : retry opening the sidebar when a demo imports its traces (#31952)

allozaur committed 4 hours ago

Verified 0a94a3f2

ui: make the app feel native on phones (#31949)

allozaur committed 4 hours ago

Verified 610f7ab4

feat: external PLE file loading (#31185)

pwilkin committed 5 hours ago

Verified 5f490d7b

Windows unbuffered model load (#26014)

JTischbein committed 6 hours ago

Verified 9d340ec3

server : functionality test for new spec dec (#31027)

nv-vankit committed 6 hours ago

Verified 359dc585

ggml-zdnn: accept non-zdnn tensors, invalidate compute tensors after every op (#31824)

taronaeo committed 10 hours ago

Verified a4d4f2fd

server : fix response_format json_schema when schema is not nested in the json_schema wrapper (#28697)

linsen458-spec committed 11 hours ago

Verified 1945793e

ggml : implement the backward pass for TRI (#30712)

losewayy committed 11 hours ago

Verified 0eebb6ed

tests : keep dummy ffn.act scales away from zero (#31940)

ggerganov committed 11 hours ago

Verified d628a98f

mtp: read the target hidden states from a device buffer (#31927)

am17an committed 11 hours ago

Verified a8fad21e

ui: fold, preview and rearrange the sidebar's sections (#31936)

allozaur committed 11 hours ago

Verified b39680c4

ggml-backend: don't upload inputs that are not read (#31419)

ArshM17-NV committed 11 hours ago

Verified 806b3c2b

models : Attach Granite NVFP4 weight scales (#31449)

jjperry69 committed 12 hours ago

Verified 0c6dca2f

ci : Run gfx1151 ROCm job on Strix Halo Devlab runners (#31567)

iswaryaalex committed 12 hours ago

Verified fc0be861

metal : prevent out-of-bounds writes to the Metal FA output buffer (#31662)

realMoana committed 12 hours ago

Verified e01c9f18

metal : make ggml_tri select instead of scaling by the comparison (#30711)

losewayy committed 12 hours ago

Verified 4b46d965

opencl : include climits (#31935)

CISC committed 13 hours ago

Verified f3178366

mtmd: Pocket TTS - add generic support for temperature set via API (#31655)

rjtokenring committed 14 hours ago

Verified 2d99adb1

RPC: Improve loading times (#29413)

am17an committed 14 hours ago

Verified cf92c132

ggml : support YaRN truncation config on openvino, sycl, vulkan, webgpu, rpc (#30802)

tokinasin committed 14 hours ago

Verified 9f0d80cf

sycl: cache device-id (#31126)

Titaniumtown committed 15 hours ago

Verified a0ad695c

tests: bound the test_rope tolerance by the f32 angle resolution (#31874)

ServeurpersoCom committed 15 hours ago

Verified 6590d25a

opencl: add f32 MUL_MAT_ID (#31783)

lhez committed 16 hours ago

Verified 53eb6694

ggml-zdnn: fix release binary not detecting backend lib (#31905)

taronaeo committed 16 hours ago

Verified d78f4255

mtmd: refactoring: remove clip.h, consolidate code into mtmd (#31865)

ngxson committed 1 day ago

Verified 7d7f0287

Older