ggml-org/llama.cpp

Pull Requests Commits

server: skip the service tests on Windows, Python has no AF_UNIX

ServeurpersoCom committed 22 hours ago

c991b33a

app: unbuffer the detached service output, MSVC rejects _IOLBF

ServeurpersoCom committed 22 hours ago

5872ecf5

server: hold a file lock for the service instead of trusting its pid

ServeurpersoCom committed 22 hours ago

8df2847d

server: build with path.string() and type the service test helpers

ServeurpersoCom committed 22 hours ago

0b287038

server: add service mode, control endpoint via .sock

ngxson committed 1 day ago

66798701

server: cap shared decision prefixes at the first output (#30710)

aiqubits committed 1 day ago

Verified f3586e66

server : fix /v1/responses endpoint to accept extra_params (#30959)

Dgarc359 committed 1 day ago

Verified 2871678a

mtmd : resize Gemma 4 images without letterbox padding (#30922)

boshjerns committed 1 day ago

Verified 7bd18083

opencl: add fwht (#31434)

shaofeiqi committed 2 days ago

Verified 0ec4027a

chat : strip the think block in the qwen3-coder with reasoning_format none (#31380)

tdakhran committed 2 days ago

Verified 1052f7e1

sycl: fix f32 to q2_0/q4_0/q5_0 rounding in CPY (#31120)

Torchit1 committed 2 days ago

Verified 8956c769

rpc : delete fs_path_to_utf8() (#31424)

angt committed 2 days ago

Verified 5047748f

opencl: add initial iq1 support (#31341)

shaofeiqi committed 2 days ago

Verified 792fcf4e

common : delete fs_path_to_utf8() (#31423)

angt committed 2 days ago

Verified a34de861

common : drop the Windows argv reconstruction (#31417)

angt committed 2 days ago

Verified aa588a34

quant : allow quantization for LFM3-Audio codec decoder (#31386)

tdakhran committed 2 days ago

Verified 889f2ec9

vendor : update cpp-httplib to 0.64.0 (#31400)

cabelo committed 2 days ago

Verified 1ab23c33

ci : add missing ccache-clear to windows behavior check (#31413)

CISC committed 2 days ago

Verified 27b8464e

cmake : embed an app manifest in all Windows exes (#31329)

angt committed 2 days ago

Verified d63fbfe7

ci : extract most apple jobs into daily workflow (#31408)

danbev committed 2 days ago

Verified e216f4ec

gemma3n, gemma4 : create token_embd before the tied output (#31394)

m4zar committed 2 days ago

Verified 344846b4

kv-cache : keep rotation constants on the GPU (#31073)

anton-iermolenko committed 2 days ago

Verified fb599727

memory: refuse rollbacks past the snapshots of the last ubatch (#31368)

ServeurpersoCom committed 2 days ago

Verified ab3768cd

opencl: add initial q2_0 support (#31340)

shaofeiqi committed 3 days ago

Verified e0f142d5

common : end reasoning budget forcing on an end tag from the prompt (#31374)

tdakhran committed 3 days ago

Verified 3c809f44

sycl: FLASH_ATTN_EXT failures (#30761)

mctylr-gh committed 3 days ago

Verified 36a0e2ff

ggml : add batched tensor copy API (#30881)

ggerganov committed 3 days ago

Verified 418f905a

ROCm/HIP: add RDNA4_M family for gfx1170/gfx1171 (#31373)

slojosic-amd committed 3 days ago

Verified eb4e244b

Quant: OCP FP8 E4M3 support (#27512)

ORippler committed 3 days ago

Verified 68d20b88

skills: specify rule about ggml_build_forward_expand (#31375)

ngxson committed 3 days ago

Verified 0893ec98

Older