[CPU] Add PagedAttention Support for Sink Input and decode phases Sliding Window #33035
liubo-intel
force pushed
from
e25dc473
to
f42fb29c
206 days ago
liubo-intel
force pushed
from
7407cb0a
to
0a391b21
202 days ago
liubo-intel
force pushed
from
0a391b21
to
d9cd8d35
202 days ago
liubo-intel
force pushed
from
908b3830
to
0b85b40f
201 days ago
liubo-intel
force pushed
from
0b85b40f
to
69bb714d
201 days ago
liubo-intel
force pushed
from
69bb714d
to
75cfc550
201 days ago
liubo-intel
force pushed
from
75cfc550
to
e0ba7725
198 days ago
maxnick
added this to the 2026.0 milestone 194 days ago
Support PA with additional sink input
a54d4863
fix PA sliding_window issue: add sliding_window process for second token
2f823123
Add PA sliding_window testcase
55bd1379
fix CI issue
e11b7616
Apply suggestions from code review
105320cd
fix rebase confilct
730d3f88
keep PA sink logic focus on constant with shape {1, H, 1, 1} to align…
042de01a
Apply suggestions from code review
700b5b11
Apply suggestions from code review
8a5bbcde
liubo-intel
force pushed
from
e0ba7725
to
8a5bbcde
193 days ago
maxnick
approved these changes
on 2025-12-10
maxnick
merged
e4d09f99
into master 192 days ago
Login to write a write a comment.
Login via GitHub