dan_the_3rd f303889ed9 fMHA: Sync FW with xFormers (#828) ...

* fMHA: Add support for bias+dropout in FW

* Remove 'getMaximumSharedMemoryPerBlockKb'

* fix comments

---------

Co-authored-by: danthe3rd <danthe3rd>
Co-authored-by: Haicheng Wu <haichengw@nvidia.com>

2023-02-22 23:25:31 -05:00

..

epilogue

fMHA: Sync FW with xFormers (#828)

2023-02-22 23:25:31 -05:00

gemm

fMHA: Sync FW with xFormers (#828)

2023-02-22 23:25:31 -05:00

iterators

xFormer updates to fMHA FW (#773)

2023-02-08 23:00:10 -05:00

transform

fMHA: Sync FW with xFormers (#828)

2023-02-22 23:25:31 -05:00

CMakeLists.txt

New updates for 2.11 (#775)

2023-01-20 16:32:57 -05:00

debug_utils.h

fMHA: Sync FW with xFormers (#828)

2023-02-22 23:25:31 -05:00

default_fmha_grouped.h

fMHA: Sync FW with xFormers (#828)

2023-02-22 23:25:31 -05:00

fmha_grouped_problem_visitor.h

New updates for 2.11 (#775)

2023-01-20 16:32:57 -05:00

fmha_grouped.h

fMHA: Sync FW with xFormers (#828)

2023-02-22 23:25:31 -05:00

fused_multihead_attention_fixed_seqlen.cu

fMHA: Sync FW with xFormers (#828)

2023-02-22 23:25:31 -05:00

fused_multihead_attention_variable_seqlen.cu

xFormer updates to fMHA FW (#773)

2023-02-08 23:00:10 -05:00

gemm_kernel_utils.h

xFormer updates to fMHA FW (#773)

2023-02-08 23:00:10 -05:00

kernel_forward.h

fMHA: Sync FW with xFormers (#828)

2023-02-22 23:25:31 -05:00