intel/auto-round

Pull Requests Commits
ark: document sparse qtile256 experiments and results
yiliu30 committed 46 days ago
6a9906a9
ark: add focused sparse-dev benchmark modes
yiliu30 committed 46 days ago
48bd8b7f
ark: extend sparse prefill runtime for decoupled rows
yiliu30 committed 46 days ago
5f55ff30
Add sparse kernel profiling modes and dev bench path
yiliu30 committed 48 days ago
99549bc6
Checkpoint sparse Sage kernel baseline
yiliu30 committed 50 days ago
c22378a2
Add opt-in Triton CUDA-port preprocess backend
yiliu30 committed 52 days ago
cefff6f0
Add top-level reproduce guide
yiliu30 committed 52 days ago
6aecab18
Document reproduce commands for Triton-XPU sparse work
yiliu30 committed 52 days ago
827a6e4b
Add Triton-XPU sparse prefill backend
yiliu30 committed 52 days ago
cde63c16
update status
yiliu30 committed 59 days ago
b38e2b5e
[ARK] Clean sparse Python glue module
yiliu30 committed 65 days ago
62a62177
[ARK] Split sparse Sage outer wrapper
yiliu30 committed 66 days ago
b3f722c8
[ARK] Document sparse attention clean branch
yiliu30 committed 66 days ago
77ea766e
[ARK] Document current sparse attention profiling status
yiliu30 committed 66 days ago
855d12a1
add nhd integration
yiliu30 committed 66 days ago
98b05279
[ARK] Update sparse kernel status and prefetch notes
yiliu30 committed 66 days ago
924dcd61
[ARK] Use copy-based sparse Sage mainloop baseline
yiliu30 committed 66 days ago
e2d6e7c1
[ARK] Add XPU sparse kernel benchmark harness
yiliu30 committed 66 days ago
b65f12cd
[ARK] Add Wan and Flux sparse attention demos
yiliu30 committed 66 days ago
15e78c8d
[ARK] Add sparse attention design and status docs
yiliu30 committed 66 days ago
c6b58f0f
[ARK] Add sparse attention tests and benchmark
yiliu30 committed 66 days ago
9d9c77dd
[ARK] Add sparse Sage kernel and preprocess path
yiliu30 committed 66 days ago
59a8c26c
fix random rotation and update rotation doc. (#1884)
lkk12014402 committed 73 days ago
Verified 4e2ad772
fix performance regression (#1886)
wenhuach21 committed 78 days ago
Verified 30f72bda
Suppress misleading warning when detecting model type for GGUF export directories (#1887)
lvliang-intel committed 78 days ago
Verified fa48924e
[ARK] Add bf16 + NHD layout support, refactor sage_dynamic_quant (#1882)
luoyu-intel committed 78 days ago
Verified 08c5d6d4
gguf format add support for mtp quantization (#1866)
n1ck-guo committed 79 days ago
Verified f7e77e03
Fix SGL CI dependency conflict (#1877)
WeiweiZhang1 committed 80 days ago
Verified 77720aea
Bump build version to v0.14.0 (#1875)
chensuyue committed 81 days ago
Verified 15e22ede
fix segmentation failure (#1871)
xin3he committed 82 days ago
Verified 22b0fa1b
Terms Of Use Privacy Policy Refund Policy Impressum