Skip to content

LLM inference in C/C++

C++ ggml

Commits on master

sycl : pass device into mul_mat_uses_onednn (#31014)
Titaniumtown committed 27 days ago
f64a498
Extend probabilistic sampling for DFlash2 drafts (#30663)
Pranesh Gonegandla committed 27 days ago
1ab09d6
vocab : fix Nemotron-3.5-Embed tokenization (#30974)
Takamichi Ohara committed 27 days ago
f9599d0
sycl: Add graph record and replay (#28725)
Todd Malsbary committed 27 days ago
7377bad
opencl: add depthwise conv2d (#30900)
kurquhar committed 28 days ago
970b51c
a08ab75
19ab5a5
Fix !30793
Gaurav Garg committed 28 days ago
36cbf41
glm5 : optimize inputs (#30932)
Georgi Gerganov committed 28 days ago
5c19a5a
d9ee007
ci : disable slow WebGPU tests (#30963)
Georgi Gerganov committed 28 days ago
5f5acfe
7170cad
vocab : add clean_spaces GGUF key (#30940)
tkran2 committed 28 days ago
d88040c
bf039c8
e3797b1
webgpu : add Q2_0 support (#30914)
Chanho Lee committed 28 days ago
0b1281a
dflash : load optional FC input scale (#30945)
Hrishith Thadicherla committed 28 days ago
4912ac2
ui: Add "Show tool call in progress" setting (#30681)
Aleksander Grygier committed 28 days ago
84746c5
lora : support prism.hadamard folded weights (#30910)
Hoang Vien Duy committed 28 days ago
0682a57