mirror of
https://github.com/saymrwulf/onnxruntime.git
synced 2026-05-18 21:21:17 +00:00
2ee822d483
### Description <!-- Describe your changes. --> 1. upgrade cutlass to 3.0 that containing attn_bias support. 2. extend Attention/MHA to use memory efficient attention when rel_pos_bias with [1, num_head, s, s*] and 1d mask with [2 * batch_size + 1] are present. new mask format introduction: MASK_1D_KEY_SEQ_LEN_START, [3 * batch_size + 2] with [key_len[0], ..., key_len[batch_size - 1], query_start[0], ..., query_start[batch_size - 1], query_end[batch_size - 1], key_start[0], ..., key_start[batch_size - 1], key_end[batch_size - 1]] e.g 2D mask with [[1, 1, 1, 0, 0, 0], [1, 1, 1, 1, 1, 0]] converts to this 1D mask is [3, 5, 0, 6, 12, 0, 6, 12] ### Motivation and Context <!-- - Why is this change required? What problem does it solve? - If it fixes an open issue, please link to the issue here. --> It potentially benefits tnlrv6 and t5(encoder) --------- Co-authored-by: Ubuntu <wy@v100-2.0cdb2e52twzevn1i4fi45bylyg.jx.internal.cloudapp.net> Co-authored-by: Kunal Vaishnavi <kvaishnavi@microsoft.com> Co-authored-by: Kunal Vaishnavi <kvaishnavi@microsoft.com@orttrainingdev7.d32nl1ml4oruzj4qz3bqlggovf.px.internal.cloudapp.net>
14 lines
411 B
CMake
14 lines
411 B
CMake
if (onnxruntime_USE_FLASH_ATTENTION)
|
|
include(FetchContent)
|
|
FetchContent_Declare(
|
|
cutlass
|
|
URL ${DEP_URL_cutlass}
|
|
URL_HASH SHA1=${DEP_SHA1_cutlass}
|
|
PATCH_COMMAND ${Patch_EXECUTABLE} --binary --ignore-whitespace -p1 < ${PROJECT_SOURCE_DIR}/patches/cutlass/cutlass.patch
|
|
)
|
|
|
|
FetchContent_GetProperties(cutlass)
|
|
if(NOT cutlass_POPULATED)
|
|
FetchContent_Populate(cutlass)
|
|
endif()
|
|
endif()
|