ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-06-29 11:16:59 +00:00
Code Issues Packages Projects Releases Wiki Activity

Commit Graph

  • 4e7ecd7f40 [Conv] Add NumGroupsToMerge to BwdWeight type string (#4271) assistant-librarian[bot] 2026-02-11 10:07:53 +01:00
  • 63dcefffc3 WIP: v4 tile distribution working Damien Lejeune 2026-02-10 13:55:07 +00:00
  • d06f35027a [rocm-libraries] ROCm/rocm-libraries#4354 (commit d41f08a) Cong Ma 2026-02-11 07:05:46 +00:00
  • 15635d4b11 [CK TILE] fix numerical errors of preshuffle_b (#4354) Cong Ma 2026-02-11 00:04:44 -07:00
  • cdba8b787c [CK TILE] fix numerical errors of preshuffle_b (#4354) Cong Ma 2026-02-11 00:04:44 -07:00
  • 807efa703a [rocm-libraries] ROCm/rocm-libraries#4274 (commit 7c380df) Thomas Ning 2026-02-11 05:52:42 +00:00
  • 39eaa5015d Add padding to cshuffle epilogue to avoid bank conflict (#4274) assistant-librarian[bot] 2026-02-10 22:52:00 -07:00
  • 844b598330 Add padding to cshuffle epilogue to avoid bank conflict (#4274) assistant-librarian[bot] 2026-02-10 22:52:00 -07:00
  • 6d6ee8f023 [rocm-libraries] ROCm/rocm-libraries#4457 (commit 258a459) Bartłomiej Kocot 2026-02-11 01:52:59 +00:00
  • 8526bd2812 [CK][CK Tile] Temporary disable grouped conv fwd tile comp async instances (#4457) Bartłomiej Kocot 2026-02-11 02:52:14 +01:00
  • 7ba919330c [CK][CK Tile] Temporary disable grouped conv fwd tile comp async instances (#4457) Bartłomiej Kocot 2026-02-11 02:52:14 +01:00
  • 9c94c2294a [rocm-libraries] ROCm/rocm-libraries#4460 (commit ba5ef82) Joseph Macaranas 2026-02-10 23:11:31 +00:00
  • 682e466cf4 [Azure External CI] Disable Azure CI on rocm-libraries (#4460) Joseph Macaranas 2026-02-10 18:02:22 -05:00
  • 88e581438e [Azure External CI] Disable Azure CI on rocm-libraries (#4460) Joseph Macaranas 2026-02-10 18:02:22 -05:00
  • eefadf7afd first working implementation for bwd_data_multiple_d_xdl_v1 Kevin Abraham 2026-02-10 21:59:08 +00:00
  • 1af75d290e [rocm-libraries] ROCm/rocm-libraries#4277 (commit 4348901) John Shumway 2026-02-10 21:27:27 +00:00
  • 201fec5e8a Add a README.md file to ck/library/util (#4277) assistant-librarian[bot] 2026-02-10 21:26:45 +00:00
  • d8a911202a Add a README.md file to ck/library/util (#4277) assistant-librarian[bot] 2026-02-10 21:26:45 +00:00
  • 72d188cd74 Merge branch 'develop_deprecated' into ckTileEnginePooling2 ckTileEnginePooling2 Aleksander Dudek 2026-02-10 18:49:23 +00:00
  • 9bfcce5566 fix formating Aleksander Dudek 2026-02-10 18:45:06 +00:00
  • d546ec0a53 [rocm-libraries] ROCm/rocm-libraries#4269 (commit 209f62f) Randy Spaulding 2026-02-10 18:38:21 +00:00
  • 8b78b47f84 Adapt parser to monorepo (#4269) assistant-librarian[bot] 2026-02-10 18:37:40 +00:00
  • 4a42d6e2ba Adapt parser to monorepo (#4269) assistant-librarian[bot] 2026-02-10 18:37:40 +00:00
  • 40cec769ce [rocm-libraries] ROCm/rocm-libraries#4266 (commit 1d8094d) Johannes Graner 2026-02-10 16:58:04 +00:00
  • 333816f842 [CK Conv] Add bwd weight instance for large-k shape (#4266) assistant-librarian[bot] 2026-02-10 16:56:59 +00:00
  • 73c1257b0e [CK Conv] Add bwd weight instance for large-k shape (#4266) assistant-librarian[bot] 2026-02-10 16:56:59 +00:00
  • afdd6a84a7 WIP: Double buffer implementation. Ville Pietilä 2026-02-10 10:03:41 -05:00
  • b41bfece83 [rocm-libraries] ROCm/rocm-libraries#4268 (commit d2fca53) Erwin Terpstra 2026-02-10 13:59:03 +00:00
  • f6bb48458d [CK_TILE]: PreshuffleB + PreshuffleBQuant for ABQuant pipeline (#4268) assistant-librarian[bot] 2026-02-10 06:57:55 -07:00
  • 2073f0a73b [CK_TILE]: PreshuffleB + PreshuffleBQuant for ABQuant pipeline (#4268) assistant-librarian[bot] 2026-02-10 06:57:55 -07:00
  • 2c2125f73e ckTileEngine pooling Aleksander Dudek 2026-02-10 12:50:42 +00:00
  • d5acfd8d52 [rocm-libraries] ROCm/rocm-libraries#4451 (commit 091bf0f) Yi DING 2026-02-10 12:42:19 +00:00
  • 1ac61a54c9 [CK_TILE] Blockscale Gemm Fix Multi-Arch Compilation (#4451) Yi DING 2026-02-10 20:41:09 +08:00
  • 824af07002 [CK_TILE] Blockscale Gemm Fix Multi-Arch Compilation (#4451) Yi DING 2026-02-10 20:41:09 +08:00
  • 59cbe19c83 More documentation. Ville Pietilä 2026-02-10 07:34:27 -05:00
  • 7c728adb57 Add V4: remove gemm pipeline, combine gemm/normalization Damien Lejeune 2026-02-10 10:39:49 +00:00
  • 6c45f722e7 Compute GEMM and normalize in one pass: MHV v3 Damien Lejeune 2026-02-10 10:35:10 +00:00
  • c9504f2c27 Update profiling doc. Ville Pietilä 2026-02-10 04:45:54 -05:00
  • 71f1333eb4 Fix types for intermediate values Matti Eskelinen 2026-02-10 09:18:48 +00:00
  • 3757000c34 ComputeDataType double fails, error out for now Matti Eskelinen 2026-02-10 08:00:46 +00:00
  • 6a6cd05dbb [rocm-libraries] ROCm/rocm-libraries#3090 (commit 728d3a3) dependabot[bot] 2026-02-10 07:08:05 +00:00
  • d61393a714 Bump fonttools from 4.57.0 to 4.61.0 in /projects/composablekernel/docs/sphinx (#3090) dependabot[bot] 2026-02-10 07:07:06 +00:00
  • 50387dd79c Bump fonttools from 4.57.0 to 4.61.0 in /projects/composablekernel/docs/sphinx (#3090) dependabot[bot] 2026-02-10 07:07:06 +00:00
  • 06ad66b3e4 [rocm-libraries] ROCm/rocm-libraries#4265 (commit 0f9b3b0) Aviral Goel 2026-02-10 03:00:40 +00:00
  • a703c54319 [CK Tools] Auto-enable unbuffered output for Python commands (#4265) assistant-librarian[bot] 2026-02-10 02:59:58 +00:00
  • 87fd831d43 [CK Tools] Auto-enable unbuffered output for Python commands (#4265) assistant-librarian[bot] 2026-02-10 02:59:58 +00:00
  • b688665d79 [rocm-libraries] ROCm/rocm-libraries#475 (commit cabe79b) dependabot[bot] 2026-02-10 02:50:35 +00:00
  • 2201d13d58 Bump pillow from 11.2.1 to 11.3.0 in /projects/composablekernel/docs/sphinx (#475) dependabot[bot] 2026-02-09 19:49:39 -07:00
  • 9ecc13eecd Bump pillow from 11.2.1 to 11.3.0 in /projects/composablekernel/docs/sphinx (#475) dependabot[bot] 2026-02-09 19:49:39 -07:00
  • 27e0a34e0f [rocm-libraries] ROCm/rocm-libraries#4406 (commit 61f9f90) Bartłomiej Kocot 2026-02-09 21:09:42 +00:00
  • 23b32f1ff8 [CK] CK Tile grouped convolution direct load (#4406) Bartłomiej Kocot 2026-02-09 22:08:57 +01:00
  • 0b9fa702ac [CK] CK Tile grouped convolution direct load (#4406) Bartłomiej Kocot 2026-02-09 22:08:57 +01:00
  • 0cafa68b6f [rocm-libraries] ROCm/rocm-libraries#4292 (commit b7f1367) Chinmay Dattanand Kuchinad 2026-02-09 20:59:55 +00:00
  • fdb1a08e6f Enable group mode (varlen) kernel generation for PyTorch integration (#4292) assistant-librarian[bot] 2026-02-09 20:58:57 +00:00
  • bb8c746cbc Enable group mode (varlen) kernel generation for PyTorch integration (#4292) assistant-librarian[bot] 2026-02-09 20:58:57 +00:00
  • b986ad8b79 Extend SplitK support to ABQuantGrouped mode ck/aviralgoel/abquant_splitk AviralGoelAMD 2026-02-09 14:19:31 -06:00
  • f2a555dac7 Align the masking logic in HstuCrossAttentionBlockMask with pytorch mask_v2 scripts Qianfeng Zhang 2026-02-09 15:55:13 +00:00
  • ea6363ad78 [rocm-libraries] ROCm/rocm-libraries#4399 (commit 331512e) Bartłomiej Kocot 2026-02-09 15:37:36 +00:00
  • 784a03af29 [CK] Fix grouped conv fwd transform for merged groups (#4399) Bartłomiej Kocot 2026-02-09 16:36:52 +01:00
  • 72016e355e [CK] Fix grouped conv fwd transform for merged groups (#4399) Bartłomiej Kocot 2026-02-09 16:36:52 +01:00
  • e16789b609 [rocm-libraries] ROCm/rocm-libraries#4373 (commit 1c29275) Eiden Yoshida 2026-02-09 15:25:01 +00:00
  • 0a9935043a [CK] MICI: Disable failure pattern checking (#4373) Eiden Yoshida 2026-02-09 10:23:47 -05:00
  • 02e6550609 [CK] MICI: Disable failure pattern checking (#4373) Eiden Yoshida 2026-02-09 10:23:47 -05:00
  • 0766752704 Refactoring the normalization operation Damien Lejeune 2026-02-09 13:55:54 +00:00
  • 8b59a1e192 Improve parallelism in the benchmark Damien Lejeune 2026-02-09 13:16:21 +00:00
  • d32bdb1412 Merge branch 'vpietila/retina-net-fwd-convs' into vpietila/retina-net-training-perf Ville Pietilä 2026-02-09 06:54:16 -05:00
  • b722492a30 Add BF16 example. Ville Pietilä 2026-02-09 06:36:20 -05:00
  • 95e46fdc79 [Compiler] Addressing a couple of more lifetime warnings amd/dev/janplehr/fix/compiler-lifetime-warning JP Lehr 2026-02-09 05:07:15 -06:00
  • 240edd0702 Add check for reference being doubly stochastic Matti Eskelinen 2026-02-09 09:43:40 +00:00
  • 60d1ec34a9 Improve benchmarking scripts. vpietila/retina-net-fwd-convs Ville Pietilä 2026-02-09 04:34:27 -05:00
  • 6f8b9548b5 Use kIsCrossAttention as Problem attribute to replace using is_cross_attention as kernel argument Qianfeng Zhang 2026-02-09 09:02:17 +00:00
  • 5b3e527c88 [rocm-libraries] ROCm/rocm-libraries#4280 (commit b7de1e1) kensclin 2026-02-09 03:55:52 +00:00
  • 4304c2c38e [CK_TILE] Add blockscale GEMM support for EightWarps on gfx950 (#4280) assistant-librarian[bot] 2026-02-09 11:54:54 +08:00
  • 6c58796a52 [CK_TILE] Add blockscale GEMM support for EightWarps on gfx950 (#4280) assistant-librarian[bot] 2026-02-09 11:54:54 +08:00
  • 731afe535a [rocm-libraries] ROCm/rocm-libraries#4357 (commit ff3e982) jakpiase 2026-02-08 19:57:53 +00:00
  • dfa95522d3 [CK_TILE] Add support and tests for V6 pipeline in conv fwd (#4357) jakpiase 2026-02-08 20:57:14 +01:00
  • 71cc990ffd [CK_TILE] Add support and tests for V6 pipeline in conv fwd (#4357) jakpiase 2026-02-08 20:57:14 +01:00
  • bdfa0a74c2 Update to hstu masking to separate the implementation for cross-attention and self-attention Qianfeng Zhang 2026-02-08 08:06:47 +00:00
  • 57d26db844 [rocm-libraries] ROCm/rocm-libraries#4273 (commit 591f504) Ville Pietilä 2026-02-08 11:35:56 +00:00
  • bb15392230 [CK] Add fwd conv group merging to v3 conv instances (#4273) assistant-librarian[bot] 2026-02-08 12:34:59 +01:00
  • f38cd21b9e [CK] Add fwd conv group merging to v3 conv instances (#4273) assistant-librarian[bot] 2026-02-08 12:34:59 +01:00
  • 4266f867d6 [rocm-libraries] ROCm/rocm-libraries#4381 (commit 5df3343) Emily Martins 2026-02-07 00:28:06 +00:00
  • 2a765fbbad [CK_TILE] Fix MMA concepts compiler error (#4381) Emily Martins 2026-02-06 17:26:57 -07:00
  • 5c31eeeddb [CK_TILE] Fix MMA concepts compiler error (#4381) Emily Martins 2026-02-06 17:26:57 -07:00
  • 4237aedf9a [rocm-libraries] ROCm/rocm-libraries#4335 (commit 06976b3) Aviral Goel 2026-02-07 00:15:34 +00:00
  • 01d37b171d Increase tolerance for FP16 GEMM tests to handle non-deterministic ro… (#4335) Aviral Goel 2026-02-07 04:14:28 +04:00
  • 92fbf5a880 Increase tolerance for FP16 GEMM tests to handle non-deterministic ro… (#4335) Aviral Goel 2026-02-07 04:14:28 +04:00
  • d2f1541976 [rocm-libraries] ROCm/rocm-libraries#4300 (commit 07e9d56) spolifroni-amd 2026-02-07 00:11:11 +00:00
  • 0b4203702c [CK] add inter/intrawave scheduling concept doc (#4300) assistant-librarian[bot] 2026-02-06 16:10:23 -08:00
  • a62115aad1 [CK] add inter/intrawave scheduling concept doc (#4300) assistant-librarian[bot] 2026-02-06 16:10:23 -08:00
  • 984a3d1828 [rocm-libraries] ROCm/rocm-libraries#4372 (commit 738ffd7) Enrico Degregori 2026-02-07 00:09:58 +00:00
  • f18a97a1f2 [CK] Workaround blockscale wp test failure (#4372) Enrico Degregori 2026-02-07 01:09:08 +01:00
  • 442c3097ee [CK] Workaround blockscale wp test failure (#4372) Enrico Degregori 2026-02-07 01:09:08 +01:00
  • c7298e57c0 remove some old files samremes/ck_tile_mx_gemm Sami Remes 2026-02-06 18:37:34 +00:00
  • 457474ed90 use stricter tolerance Sami Remes 2026-02-06 18:28:19 +00:00
  • 1622674c9e use persistent Sami Remes 2026-02-06 18:21:34 +00:00
  • 1ddb38f098 [rocm-libraries] ROCm/rocm-libraries#4375 (commit 45b616b) Illia Silin 2026-02-06 18:18:14 +00:00
  • b5d4754abf [CK] fix path for build filter (#4375) Illia Silin 2026-02-06 10:17:02 -08:00
  • 8cd3f55a72 [CK] fix path for build filter (#4375) Illia Silin 2026-02-06 10:17:02 -08:00
  • dc4366a876 add main include file Sami Remes 2026-02-06 18:12:54 +00:00