Commit Graph

Select branches

Hide Pull Requests

main

v4.2.1

Mono Color

• 8c1bf9b784 Bump CUTLASS Python container version (#672) Jack Kosaian 2022-10-22 21:09:39 -04:00
• 7d0dd6706e Remove excessive includes from examples/41_multi_head_attention (#669) Yuriy Chernyshov 2022-10-22 04:23:15 +02:00
• 9b47403b2d Add missing CUTLASS_HOST_DEVICE (#671) hlu1 2022-10-21 19:20:38 -07:00
• 4db6a6140e ex42: Fused MHA imported from xFormers (#662) dan_the_3rd 2022-10-17 16:49:33 +02:00
• 3bf95e90c2 Update labeler.yml Matthew Nicely 2022-10-13 08:03:28 -04:00
• 75fed7493e Update labeler.yml Matthew Nicely 2022-10-13 08:01:21 -04:00
• 98b73fc95d Update labeler.yml Matthew Nicely 2022-10-13 07:55:33 -04:00
• 4990e3686d Update labeler.yml Matthew Nicely 2022-10-13 07:52:38 -04:00
• 4b7365388c Update labeler.yml Matthew Nicely 2022-10-13 07:32:55 -04:00
• 0d8405588d Update labeler.yml Matthew Nicely 2022-10-12 15:32:38 -04:00
• cb539dab78 Correct typos in comments (#639) Alexander Freudenberg 2022-10-01 04:51:30 +02:00
• dadc881a96 Bug fix for gemm broadcast (#650) Ying Zhang 2022-09-30 07:00:38 -07:00
• f3eea3a4d7 Create labeler.yml Matthew Nicely 2022-09-29 15:08:44 -04:00
• cd37e82492 change unused class member to local var (#646) Wenzhuo Liu 2022-09-29 11:52:35 +08:00
• 48a9ea223a Fix release version in the citation (#638) ANIKET SHIVAM 2022-09-22 07:58:45 -07:00
• 7a458f00a6 fix(permute.h): incorrect comment in Tensor5DPermute20314 (#637) Wenzhuo Liu 2022-09-22 21:21:13 +08:00
• 97bff52e8c add two missing files (#636) Haicheng Wu 2022-09-21 15:42:42 -04:00
• 9f2e3faa69 fix call of GELU_Taylor in LinearCombinationGeneric (#634) Tianqi Zhang (张天启) 2022-09-21 09:00:55 +08:00
• a821280dc7 Gemm broadcast (#632) Ying Zhang 2022-09-20 07:37:12 -07:00
• f73374a1eb fix:comment typo in example 23 (#633) Wenzhuo Liu 2022-09-19 21:54:14 +08:00
• faab7536fc add comment (#628) Yujia Zhai 2022-09-17 18:40:30 -07:00
• fc9ebc645b CUTLASS 2.10 bug fixes and minor updates. (#626) Andrew Kerr 2022-09-15 16:20:33 -04:00
• 2cc2c7ba1f Add set_k_partition function (#624) alexfreudenberg 2022-09-14 04:34:20 +02:00
• 50ceed7154 Minor README fix (#623) ANIKET SHIVAM 2022-09-12 19:40:25 -07:00
• e773429f7e CUTLASS 2.10 updates (#622) ANIKET SHIVAM 2022-09-12 18:26:30 -07:00
• beae168f90 fix broken link (#620) Yujia Zhai 2022-09-06 13:32:44 -07:00
• f29d8f7ca9 Include vector in base_grouped.h (#618) Jack Kosaian 2022-09-06 13:21:23 -04:00
• b1d3f9b2fd upstream internal updates (#616) Yujia Zhai 2022-09-04 20:05:09 -07:00
• b72cbf957d CUTLASS 2.10 (#615) ANIKET SHIVAM 2022-09-03 15:48:46 -07:00
• ca23ff7924 Fixed typo in class name (#608) Cliff Burdick 2022-08-29 17:51:52 -07:00
• 1c3d400b14 Added value_type trait to complex to make it an easier drop-in replacement for std::complex. (#607) Cliff Burdick 2022-08-27 22:12:40 -07:00
• abafbf2afd Missing comma in trmm header (#604) Cliff Burdick 2022-08-25 13:07:33 -07:00
• 536b20763e Fixed typo in profiler README (#603) Cliff Burdick 2022-08-24 19:55:13 -06:00
• 497b499d9d Add residual support for shmem staging iterator used in back-to-back GEMM fusion. This allows support of problem_size_0_n that is not multiple of 32. (#590) Haicheng Wu 2022-08-15 11:19:24 -04:00
• e66bfcb1f8 Fix for #596 (typo in example 03) (#597) Jack Kosaian 2022-08-09 09:58:36 -04:00
• 1617685a77 fix: fix types in example 06 (#587) Michaël Benesty 2022-07-29 18:46:06 +02:00
• 25ebf15d02 Ensure all arch::Mma specializations have ElementC set (#576) dan_the_3rd 2022-07-23 05:53:03 +02:00
• 5d05808072 fix gather example (#574) Shang Zhang 2022-07-19 13:18:17 -07:00
• 0b8cacd6f1 Remove redundant <fstream> includes (#563) Ivan Komarov 2022-07-19 22:23:54 +03:00
• e7a61c761a fix race condition when h < stride_h or w < stride_w (#562) Haicheng Wu 2022-07-12 16:37:08 -04:00
• fb379eaa5b epilogue leaky relu support ScaleType (#564) seventh 2022-07-12 05:30:55 +08:00
• 8a766804ad Fix doc in testbed_gemm_with_broadcast (#559) Jacob He 2022-07-07 06:56:16 -07:00
• 1eb6355182 [activation] tanh (#550) Bing Xu 2022-07-02 05:00:45 -07:00
• 04a9777b87 Softmax (#546) Yujia Zhai 2022-07-01 22:19:18 -07:00
• e45e773436 Update linear_combination_generic.h (#472) Haicheng Wu 2022-06-28 07:29:38 -04:00
• dae6b6893b Update CHANGELOG.md Haicheng Wu 2022-06-27 23:30:49 -04:00
• ba18ea9c32 Update README.md Haicheng Wu 2022-06-27 23:25:26 -04:00
• 9ab9110168 add leaky relu (#542) Haicheng Wu 2022-06-26 10:07:50 -04:00
• e5d4669f16 Update CHANGELOG.md (#543) Jinze (Richard) Xue 2022-06-25 10:23:49 -07:00
• 94f01f19d5 Add implicit gemm perf Haicheng Wu 2022-06-23 22:47:11 -04:00
• fa56763c25 Fix occupancy calculation for grouped GEMM (#532) Jack Kosaian 2022-06-18 19:53:59 -04:00
• 25e26a6e51 fix bugs in linear_combination_generic.h missing include cutlass/epilogue/thread/scale_type.h (#531) LiuWei 2022-06-18 11:35:14 +08:00
• f248e9bdb4 Create CITATION.cff Haicheng Wu 2022-06-07 21:25:16 -04:00
• dceefe4f64 Increment stride correctly in warp iterator. (#516) Pei Sun 2022-06-06 09:33:36 -07:00
• c3881d097e Fix a comment about LDSM layout. (#514) Pei Sun 2022-06-04 20:04:00 -07:00
• a29dfb1c63 Fix a bug to increment stride tile correctly (#503) Pei Sun 2022-06-03 19:54:52 -07:00
• 0abaac84ea [examples] Fix typos in SYRK and TRMM examples (#507) Jack Kosaian 2022-06-03 22:52:41 -04:00
• 858c735856 Update gather_scatter_fusion.cu Haicheng Wu 2022-05-18 13:15:25 -04:00
• d6f58b2d14 Update functionality.md Haicheng Wu 2022-05-11 09:34:24 -04:00
• c4cf0dad82 Fix init-self compiler warnings (#493) Mike Iovine 2022-05-10 21:35:28 -07:00
• 57551902d0 Update functionality.md Haicheng Wu 2022-05-11 00:01:19 -04:00
• 1604ebaf10 Update generator.py Haicheng Wu 2022-05-08 21:47:15 -04:00
• 6023038bae add verification of the reduction tensor (#489) Haicheng Wu 2022-05-06 13:24:51 -04:00
• ddd8f9cf41 update float < int32_t * 4 (#488) TonyZhao 2022-05-05 01:36:05 +08:00
• ec2b4fd85d b2b bias vector support (#482) Haicheng Wu 2022-04-30 07:16:15 -04:00
• 86ce09aed1 2.9 fixes for nvrtc (#480) Stepan Tezyunichev 2022-04-29 16:06:52 +03:00
• 21c1fa3849 add .github (#479) Haicheng Wu 2022-04-28 15:36:59 -04:00
• 8c339ac039 Fix compilation in clang (#478) Janusz Lisiecki 2022-04-28 20:22:06 +02:00
• e49f690fd7 Update linear_combination_generic.h Haicheng Wu 2022-04-28 14:04:53 -04:00
• 96dad61a75 Update CHANGELOG.md Haicheng Wu 2022-04-28 10:52:10 -04:00
• cc2ea4c3fc Update README.md Haicheng Wu 2022-04-28 10:50:11 -04:00
• a0de301283 Used relative paths for includes (#477) Andrew Kerr 2022-04-27 15:04:23 -04:00
• 319a389f42 Update CMakeLists.txt (#473) Haicheng Wu 2022-04-27 10:02:26 -04:00
• 71def2f084 Use platform:: instead of std::abs and std::conditional (#452) Stepan Tezyunichev 2022-04-25 21:40:22 +03:00
• 70f3ba57f5 Fix typo in shared memory layout description (#471) Masahiro Masuda 2022-04-25 07:32:13 +09:00
• dd77fadc70 Remove redundant offset def and init in shared_load_iterator.h (#456) Fujun Han 2022-04-25 04:31:00 +08:00
• be4578d517 Fixed template struct/class mismatch (#453) Stepan Tezyunichev 2022-04-24 23:30:21 +03:00
• d7b499deff Fix CUDA_PERROR_EXIT and print failing expression (#446) Andrei Alexandrescu 2022-04-24 16:29:43 -04:00
• 310ed81ac3 fix description in example 12. (#444) Exusial 2022-04-25 04:29:06 +08:00
• 4c0d6e1eb4 [BUGFIX]: Force unroll a loop that doesn't have compilation constant (#441) Fujun Han 2022-04-25 04:28:32 +08:00
• 167ac54c65 Fix link to Python example (#469) Jack Kosaian 2022-04-23 15:37:38 -04:00
• 12f4108ac2 CUTLASS 2.9 (#468) Andrew Kerr 2022-04-23 15:02:38 -04:00
• dd571f0edb [style] fix code indentation (#449) Feng Shijie 2022-04-04 09:13:17 +08:00
• 6d0d265047 Update PUBLICATIONS.md (#447) Jianyu Huang 2022-04-03 18:03:28 -07:00
• f11fa975a5 Update PUBLICATIONS.md Haicheng Wu 2022-03-23 21:04:43 -04:00
• 0e71d9b450 Transposed conv2d and wgrad split k examples (#413) Masahiro Masuda 2022-03-24 03:52:54 +09:00
• eb0d4c9213 [library] pass pointer of arguments to get_host_workspace_size() in gemm_universal() (#412) Minmin Sun (孙敏敏) 2022-03-23 00:36:34 +08:00
• bc45e2c023 fixed datatype error of numeric_limit for uint1b_t (#419) Haojin Yang 2022-03-22 17:30:30 +01:00
• 095cbba57c Example 23 - Passing correct alpha and beta values with --parallel-split-k (#424) Yang Chen 2022-03-22 09:27:34 -07:00
• 8f1fe7a132 Fix separate compilation -dc (#433) Janusz Lisiecki 2022-03-22 17:21:18 +01:00
• 3ab1eacf09 Fix typo in profiler examples (#437) Yuanqiang Liu 2022-03-22 00:00:13 +08:00
• cd39c75e25 Fix typo in docs, code comments (#429) Feng Shijie 2022-03-16 09:54:36 +08:00
• b2e1e97cb1 Update PUBLICATIONS.md Haicheng Wu 2022-03-01 22:37:18 -05:00
• 96a11a1ef3 Removed trivial copy constructors on parameter classes to enable devi… (#366) HouQiming 2022-03-01 10:34:02 +08:00
• e96f00586c Make cutlass::gemm::device::GemmArray usable (#295) Ivan Komarov 2022-02-18 04:01:05 +03:00
• 3cfa5db2a2 Actually use float accumulation in gemm_f16t_f16t_f16t_wmma_tensor_op… (#407) Jongsoo Park 2022-02-16 06:53:21 -08:00
• 1db6971a8d Remove unused gemm_k_iterations in GemmKernel::Params (#406) Jongsoo Park 2022-02-16 06:52:45 -08:00
• b954127297 Update PUBLICATIONS.md Haicheng Wu 2022-02-14 16:54:32 -05:00
• d0d941efc7 [hardswish] correct implmentation (#403) Bing Xu 2022-02-09 11:28:53 -08:00
• 8a951b2940 Enable convolution with fused epilogue for Volta Tensor Cores (#402) Andrew Kerr 2022-01-30 23:24:50 -05:00