Skip to content
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

[MHLO] Support for dynamic shape in basic op conversion by introducing CHLO dialect #1123

Merged

Conversation

Vremold
Copy link
Collaborator

@Vremold Vremold commented Jul 29, 2022

See RFC #999
This is a revision of PR #1116 . The main change is switching to use CHLO dialect to support dynamic shape when lowering basic op to MHLO dialect.

Co-authored-by: Bairen Yi yibairen.byron@bytedance.com
Co-authored-by: Jiawei Wu xremold@gmail.com
Co-authored-by: Tianyou Guo tianyou.gty@alibaba-inc.com
Co-authored-by: Xu Yan yancey.yx@alibaba-inc.com
Co-authored-by: Ziheng Jiang ziheng.jiang@bytedance.com

@Vremold Vremold force-pushed the upstream-03-init-mhlo-basic-op-conversion-dynamic branch from 2ceaca5 to b8b7d22 Compare July 31, 2022 07:39
@Vremold
Copy link
Collaborator Author

Vremold commented Jul 31, 2022

Ready for review @silvasean @powderluv @fortianyou

Copy link
Collaborator

@Yancey1989 Yancey1989 left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM with a tiny comment.

lib/Conversion/TorchToMhlo/BasicOp.cpp Show resolved Hide resolved
Copy link
Collaborator

@tanyokwok tanyokwok left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

LGTM

Copy link
Collaborator

@ZihengJiang ZihengJiang left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Great work Jiawei!

wujiawei.jw added 4 commits August 2, 2022 09:51
…g CHLO dialect

Co-authored-by: Bairen Yi <yibairen.byron@bytedance.com>
Co-authored-by: Jiawei Wu <xremold@gmail.com>
Co-authored-by: Tianyou Guo <tianyou.gty@alibaba-inc.com>
Co-authored-by: Xu Yan <yancey.yx@alibaba-inc.com>
Co-authored-by: Ziheng Jiang <ziheng.jiang@bytedance.com>
@Vremold Vremold force-pushed the upstream-03-init-mhlo-basic-op-conversion-dynamic branch 2 times, most recently from 1b6ce47 to 24264e9 Compare August 2, 2022 02:07
@Vremold Vremold merged commit 76c9766 into llvm:main Aug 2, 2022
@Vremold
Copy link
Collaborator Author

Vremold commented Aug 2, 2022

Merged. Thanks @Yancey1989 @ZihengJiang @fortianyou

@Vremold Vremold deleted the upstream-03-init-mhlo-basic-op-conversion-dynamic branch August 4, 2022 02:35
qedawkins pushed a commit to nod-ai/torch-mlir that referenced this pull request Oct 3, 2022
* disable simd for matmul when simdization dim is smaller than simd vector in MatMul operator

Signed-off-by: Alexandre Eichenberger <alexe@us.ibm.com>
Co-authored-by: Tung D. Le <tungld@gmail.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

4 participants