Skip to content

Pull requests: ROCm/composable_kernel

New pull request New
41 Open 1,424 Closed
41 Open 1,424 Closed
Author
Filter by author
Loading
Label
Filter by label
Loading
Use alt + click/return to exclude labels
or ⇧ + click/return for logical OR
Projects
Filter by project
Loading
Milestones
Filter by milestone
Loading
Reviews
Filter by reviews
No reviews Review required Approved review Changes requested
Assignee
Filter by who’s assigned
Sort
Sort by
Newest Oldest Most commented Least commented Recently updated Least recently updated Best match
Most reactions
👍 👎 😄 🎉 😕 ❤️ 🚀 👀

Pull requests list

[CK_TILE] Paged Attention use larger tile size for chunk-prefill
#1700 opened Nov 28, 2024 by poyenc Loading…
@poyenc
2
Add IsSupportedArgument() to gemm_kernel
#1698 opened Nov 27, 2024 by jakpiase Loading…
LWPCK-2429: Device grouped GEMM uses Async Memcpy
#1695 opened Nov 26, 2024 by rtmadduri Loading…
@rtmadduri
2
Fix PR1676
#1694 opened Nov 26, 2024 by aosewski • Draft
@aosewski
[Python] Add batched gemm instances parsing
#1684 opened Nov 21, 2024 by tenpercent Loading…
[BUGFix] Reorder GetSmemSize Related API
#1683 opened Nov 21, 2024 by LeiWang1999 Loading…
@LeiWang1999
1
[CK_TILE] Add 4x64x4 & 64x4x4 fp16/bf16 warp gemms
#1669 opened Nov 15, 2024 by poyenc • Draft
@poyenc
add batched_transpose implement
#1660 opened Nov 13, 2024 by fangche123 • Draft
@ThomasNing
8
Apply universal gemm to bwd_weight_cshuffle operator
#1658 opened Nov 12, 2024 by mozga-amd • Draft
4
Update ck_a8w8
#1636 opened Nov 5, 2024 by aska-0096 Loading…
A prototype of TF32 gemm
#1632 opened Nov 4, 2024 by zjing14 • Draft
[do not review] int4 scale based on jzhang's pre work noCI Disable testing on supported CI systems: math libraries CI has this feature enabled..
#1628 opened Nov 4, 2024 by mtgu0705 • Draft
[DO NOT REVIEW] Add int4 dequant with scale only supports based on ZhangJing's PR #1572 noCI Disable testing on supported CI systems: math libraries CI has this feature enabled..
#1627 opened Nov 4, 2024 by mtgu0705 • Draft
Ck tile batched gemm example
#1615 opened Oct 30, 2024 by aledudek Loading…
36
add rounding converter
#1599 opened Oct 25, 2024 by coderfeli Loading…
[GEMM] Congruous GEMM optimization enhancement New feature or request
#1590 opened Oct 21, 2024 by aska-0096 • Draft
@aska-0096
Codegen hipRTC compilation
#1579 opened Oct 16, 2024 by arai713 Loading…
Rebase the PR #1520 to ROCm repo.
#1574 opened Oct 15, 2024 by illsilin Loading…
[EXPERIMENTNAL][DO NOT MEREG] Add a prototype of F16/BF16xINT4 GEMM
#1572 opened Oct 13, 2024 by zjing14 • Draft
universal gemm bias activation example
#1569 opened Oct 12, 2024 by ltqin Loading…
1
Introduce gemm_elementwise_gemm
#1543 opened Oct 2, 2024 by mirza-halilcevic Loading…
1
Introduce gemm_softmax_gemm to codegen
#1542 opened Oct 2, 2024 by mirza-halilcevic Loading…
1
2
BF16 GEMM Stream-K bug Something isn't working
#1541 opened Oct 2, 2024 by ozturkosu • Draft
@ozturkosu
2
[WIP] add more example for permute/scatter-gather/moe/tile-reduce/fa
#1515 opened Sep 16, 2024 by carlushuang Loading…
13 of 17 tasks
@carlushuang
CMake: Enable cache if available + Prevent in-source builds
#1507 opened Sep 12, 2024 by mozga-amd • Draft
1
ProTip! Filter pull requests by the default branch with base:develop.