-
Notifications
You must be signed in to change notification settings - Fork 140
Pull requests: Tencent/hpc-ops
Author
Label
Projects
Milestones
Reviews
Assignee
Sort
Pull requests list
refactor: per-architecture operator layout with dispatch at the launch site
#90
opened Aug 26, 2026 by
qsang-nv
Loading…
chore: unify remaining copyright headers to Tencent 2026
#88
opened Aug 25, 2026 by
xueyangcs
Contributor
Loading…
fix: move pipeline barriers into dynamic SharedStorage for 128-byte alignment (#42)
#87
opened Aug 24, 2026 by
XFDG
Loading…
# docs: add layout algebra comment to make_tiled_copy_V_interleave_trans (#44)
#86
opened Aug 24, 2026 by
XFDG
Loading…
fix: fix kTaskLoopPolicy type in grouped gemm host code
#85
opened Aug 23, 2026 by
icyhearts
Loading…
feat: support head_dim=256 for SM90 BF16 decode
#84
opened Aug 20, 2026 by
Carlos779988
Loading…
4 tasks done
fix: change kHalfFltEps from 1e-20 to 5e-8 in sampler rng, aligned to torch
#65
opened Jul 13, 2026 by
LINZHI97
Loading…
feat: relax heads_per_group constraint to support 1~8 for BF16 and FP8 decode attention
#58
opened Jun 25, 2026 by
Religious-J
Contributor
Loading…
ProTip!
Updated in the last three days: updated:>2026-08-24.