"examples/offline_inference/qwen_1m.py" did not exist on "f6518b2b487724b3aa20c8b8224faba5622c4e44"
- 11 Mar, 2026 19 commits
-
-
Giancarlo Delfin authored
Signed-off-by:Giancarlo Delfin <gdelfin@inferact.ai>
-
Wentao Ye authored
Signed-off-by:yewentao256 <zhyanwentao@126.com>
-
Or Ozeri authored
Signed-off-by:
Or Ozeri <oro@il.ibm.com> Co-authored-by:
Nicolò Lucchesi <nlucches@redhat.com>
-
Hongxin Xu authored
Signed-off-by:
arlenxu <arlenxu@tencent.com> Signed-off-by:
xhx1022 <1737006628@qq.com> Co-authored-by:
-
Jhao-Ting Chen authored
Signed-off-by:
Izzy Putterman <iputterman@nvidia.com> Signed-off-by:
Jhao-Ting Chen <jhaotingc@nvidia.com> Co-authored-by:
-
Woosuk Kwon authored
Signed-off-by:Woosuk Kwon <woosuk@inferact.ai>
-
tvirolai-amd authored
Signed-off-by:Teemu Virolainen <teemu.virolainen@amd.com>
-
Wuxun Zhang authored
Signed-off-by:Zhang, Wuxun <wuxun.zhang@intel.com>
-
YiSheng5 authored
Signed-off-by:yisheng <yi.sheng@intel.com>
-
Nicolò Lucchesi authored
Signed-off-by: -
JartX authored
[Bugfix] Add Multiple of 16 block_size to triton fallback on rocm Attention to support qwen3_5 (#35923) Signed-off-by:
JartX <sagformas@epdcenter.es> Co-authored-by:
akaratza <akaratza@amd.com> Co-authored-by:
TJian <tunjian.tan@embeddedllm.com>
-
pschlan-amd authored
Signed-off-by:Patrick Schlangen <pschlan@amd.com>
-
Sladyn authored
Signed-off-by:sladynnunes <snunes@usc.edu>
-
Hongbin Guo authored
Signed-off-by:Hongbin10 <jdmjdm1998@163.com>
-
Matthew Bonanni authored
Signed-off-by:Matthew Bonanni <mbonanni@redhat.com>
-
Wentao Ye authored
Signed-off-by:
Wentao Ye <44945378+yewentao256@users.noreply.github.com> Co-authored-by:
-
Benjamin Chislett authored
Signed-off-by:Benjamin Chislett <bchislett@nvidia.com>
-
fangyuchu authored
[Bugfix] Surface exceptions from non-blocking execute_model in UniProcExecutor to avoid DP deadlocks (#35194) Signed-off-by:fangyuchu <fangyuchu@qq.com>
-
Matthew Bonanni authored
Signed-off-by:
-
- 10 Mar, 2026 14 commits
-
-
Woosuk Kwon authored
Signed-off-by: -
Nick Hill authored
Signed-off-by:Nick Hill <nickhill123@gmail.com>
-
Woosuk Kwon authored
Signed-off-by: -
Harry Mellor authored
Signed-off-by:Harry Mellor <19981378+hmellor@users.noreply.github.com>
-
Pleaplusone authored
Signed-off-by:ganyi <ygan@amd.com>
-
Matthew Bonanni authored
Signed-off-by: -
Srinivasoo7 authored
feat(kv-offload): Strategy A — StoreReusedOffloadingManager gates CPU stores on reuse frequency (#35342) Signed-off-by:
srinivas_oo7 <Sriusa4414@gmail.com> Signed-off-by: Sriusa4414@gmail.com Signed-off-by:
Srinivasoo7 <158864704+Srinivasoo7@users.noreply.github.com> Co-authored-by:
srinivas_oo7 <sklinkedin0120@gmail.com> Co-authored-by:
-
SoluMilken authored
Signed-off-by:SoluMilken <ypiheyn.imm02g@g2.nctu.edu.tw>
-
Mark McLoughlin authored
Signed-off-by:Mark McLoughlin <markmc@redhat.com>
-
Vadim Gimpelson authored
Signed-off-by:Vadim Gimpelson <vadim.gimpelson@gmail.com>
-
Nick Hill authored
Signed-off-by: -
Nick Hill authored
Signed-off-by: -
Wentao Ye authored
[Perf] Compute maxsim in worker side, reducing redundant copies, 2.7% E2E throughput improvement (#36159) Signed-off-by: -
Woosuk Kwon authored
Signed-off-by:
-
- 09 Mar, 2026 7 commits
-
-
Woosuk Kwon authored
Signed-off-by: -
Lucas Wilkinson authored
Signed-off-by:Lucas Wilkinson <lwilkins@redhat.com>
-
Matthew Bonanni authored
Signed-off-by: -
Woosuk Kwon authored
Signed-off-by: -
Woosuk Kwon authored
Signed-off-by: -
Andreas Karatzas authored
[ROCm][CI] Fix ROCm attention backend validation for head sizes, block sizes, and compute capability checks (#36292) Signed-off-by: -
Roberto L. Castro authored
[Attention][Perf][Kernel] Replace torch.cat with vectorized CUDA kernel MLA query concat - DeepSeek-V3.2 (#34917) Signed-off-by:
LopezCastroRoberto <rocastro@redhat.com> Signed-off-by:
Roberto L. Castro <38211239+LopezCastroRoberto@users.noreply.github.com>
-
