ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-06-28 18:56:59 +00:00
Code Issues Packages Projects Releases Wiki Activity
3,438 Commits 776 Branches 40 Tags

Default Branch

develop

2089713f94 · [rocm-libraries] ROCm/rocm-libraries#8227 (commit 75c30d5) · Updated 2026-06-26 12:00:58 +00:00
 ZIP  TAR.GZ

Branches

fp16_int4_scale_weight_only_impl_bak

3bec0b7232 · fixed a bug in the ckProfiler, and formatted the code · Updated 2025-01-03 06:14:36 +00:00    ROCm

1829
9
 ZIP  TAR.GZ
int4_weight_only_from_mtgu_golden

6e7218fb08 · Fixed example_gemm_xdl_bf16_pk_i4_v3 test failed issue · Updated 2025-01-01 07:48:45 +00:00    ROCm

1843
78
 ZIP  TAR.GZ
Ali_FA3_BWD

acddcee9b5 · updata fa3 bwd hasco of A32_RTZ_CasualOrNon_Bf16OrFp16 · Updated 2024-12-31 02:04:32 +00:00    ROCm

1974
27
 ZIP  TAR.GZ
mtgu/dev/deepseekv3_ab_scale

5e91d8444f · Change constant to variable · Updated 2024-12-30 11:47:30 +00:00    ROCm

1838
6
 ZIP  TAR.GZ
dev/a8w8_b_preshuffle

db84352941 · fix warnings and revert cmake and fix clang format · Updated 2024-12-30 08:24:50 +00:00    ROCm

1834
13
 ZIP  TAR.GZ
dev/merge_u8w8_onlygemm_build

b9608dc3b7 · merge; impl fp16 · Updated 2024-12-27 07:54:07 +00:00    ROCm

1834
5
 ZIP  TAR.GZ
te_bwd_v3_hd64

5081247872 · enable bias for te · Updated 2024-12-26 21:34:50 +00:00    ROCm

1974
27
 ZIP  TAR.GZ
pytorch2.6_base

4f8f655b4d · Include __gfx9__ declaration for gfx950 also · Updated 2024-12-23 18:27:53 +00:00    ROCm

1878
1
 ZIP  TAR.GZ
dev/cka8w8_tuning

2af456af18 · fix · Updated 2024-12-23 10:35:12 +00:00    ROCm

1858
45
 ZIP  TAR.GZ
ck/opt_a8w8_m32

3ea889f3b0 · add instances · Updated 2024-12-20 15:07:02 +00:00    ROCm

1837
1
 ZIP  TAR.GZ
docs/6.3.1

e24308ec91 · add requirements.txt · Updated 2024-12-19 19:36:54 +00:00    ROCm

1974
103
 ZIP  TAR.GZ
kqian/lynm+bias+quant

fd92d7eb9d · Trigged dynamic layernorm on · Updated 2024-12-16 08:14:26 +00:00    ROCm

1855
6
 ZIP  TAR.GZ
te_v3_bwd_sbhd

7cccb0a870 · Enable sbhd for customer case v3 bwd debug · Updated 2024-12-15 05:30:37 +00:00    ROCm

1974
26
 ZIP  TAR.GZ
bd_a8w8

5e8b570372 · add multiple multiple add profiler · Updated 2024-12-11 02:22:47 +00:00    ROCm

1918
1
 ZIP  TAR.GZ
yewang_qkv_sbhd

8241459e2a · change the layout to sbhd for debugging · Updated 2024-12-06 06:12:46 +00:00    ROCm

2095
50
 ZIP  TAR.GZ
gemm_bf16_sk_bug_fix

74f50fe0da · bug was discovered succesfully which is :: dst_buf_template Update not happen if DstImMemOp is atomicAdd for bf16 · Updated 2024-12-05 11:58:20 +00:00    ROCm

1982
20
 ZIP  TAR.GZ
ck_tile/fused_moe_general

593dd7ad64 · clear some code · Updated 2024-12-04 10:41:31 +00:00    ROCm

1922
49
 ZIP  TAR.GZ
ck_tile/gemm_debug_alias

99c8123f48 · change to static · Updated 2024-12-04 08:47:59 +00:00    ROCm

1880
19
 ZIP  TAR.GZ
ck_tile/gemm_opt

c275904b7d · try to fix hint · Updated 2024-12-03 06:53:32 +00:00    ROCm

1880
17
 ZIP  TAR.GZ
ck_tile/gemm_felix_test

6a07464b56 · change ways but still could not use immediate data as ds_read · Updated 2024-11-28 08:54:32 +00:00    ROCm

1912
8
 ZIP  TAR.GZ