-
Notifications
You must be signed in to change notification settings - Fork 20
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
Implement Permutation with CK #150
Conversation
3188d4b
to
bcedba6
Compare
- Add CPU implementation of elementwise permutation - permutationByCpu - Add sample of elementwise permutation GPU implementation of elementwise permutation is not done yet.
moved 01_contraction/device/common.hpp -> device/common.hpp renamed 01_contraction/common.hpp -> utils.hpp since there is already a file named common.hpp in test folder
bcedba6
to
10462aa
Compare
- implemented `hiptensorPermutation` with CK - added unit tests for `hiptensorPermutation`
10462aa
to
bf1c7ee
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
We might have to think about how to treat the __half datatype.
At some level we will have to wrap __half as _Float16 as they follow the same IEEE format.
@CongMa13 we should coordinate with CI about updating the CK package so that pre-checkin should pass |
551e784
to
790584a
Compare
Data layout of cuTenor is col_major by default. HipTensor contraction only supports row_major for now. HipTensor supports both col_major and row_major to make users be able to choose which layout they want to use. - Use CMake option `HIPTENSOR_DATA_LAYOUT_COL_MAJOR` to choose layout
790584a
to
082a966
Compare
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
LGTM!
00fac42
to
b761533
Compare
new file: include/native_types.hpp new file: include/native_types_impl.hpp new file: include/type_traits.hpp new file: include/types_ext.hpp new file: include/xfloat32.hpp
b761533
to
53d863a
Compare
No description provided.