LLM inference in C/C++
C++ ggml
Commits on master
[UT] enhance UT to show all real unsupported backends (#25234)
a6a77bc Neo Zhang committed 3 months ago
fix: increase greeting spacing on md screens (#26287)
3018a11 Aleksander Grygier committed 3 months ago
llama: move suppress_tokens handling to common/sampling (#26276)
afeebe1 Xuan-Son Nguyen committed 3 months ago
ggml-cuda : disable MMQ on devices with less than 48 KiB shared memory (#26141)
caa596a Kakaru committed 3 months ago
sycl: contiguous fast path + 32-bit index math for unary elementwise ops (#25946)
11b068d Titaniumtown committed 3 months ago
vendor: update BoringSSL to 0.20260728.0 (#26241)
e2f59ed Alessandro de Oliveira Faria (A.K.A.CABELO) committed 3 months ago
server : add trace logging for slot similarity checking (#26271)
992c325 Georgi Gerganov committed 3 months ago
conversion: fix Qwen2.5-Omni mmproj conversion regression (#26262)
e1af89a Kaben Nanlohy committed 3 months ago
add rdna3.5, and 3 to mmq configs so they can be tuned independently. (#26199)
60bccc3 Geramy Loveless committed 3 months ago
model: add NextN/MTP speculative decoding support for GLM_DSA (GLM-5.2) (#25980)
7be2c65 Satinder Grewal committed 3 months ago
ggml-webgpu: Fix some binding alias issues to support all archs, fix recurrent-state-rollback test (#25931)
bc71c24 Reese Levine committed 3 months ago
opencl: skip the Adreno KQ/KQV image kernels for multi-stream batches (#26189)
8190848 Hongqiang Wang committed 3 months ago
mtmd : add Nemotron 3 Nano Omni support (parakeet) (#22520)
7e1e28c Daniel Bevenius committed 3 months ago
server: abstract llama_memory calls to common_memory (#26221)
ee3d1b5 Xuan-Son Nguyen committed 3 months ago
ggml-cuda: add chunked SSD matmul for Mamba-2 prefill acceleration (#22675)
b62b350 Bhavik Sharda committed 3 months ago
common/chat: add specialized minimax m3 parser (#26210)
6ba5ef2 Aldehir Rojas committed 3 months ago
sycl: fix use-after-return of the SDPA scale in the oneDNN flash-attention path (#25880)
d6b61ac meatposes committed 3 months ago
server : add extra trace log for prompt similarity (#26218)
9a3bf2b Georgi Gerganov committed 3 months ago
contrib : add guideline about the "merge ready" label (#26178)
c6292cf Georgi Gerganov committed 3 months ago