This repository was archived by the owner on Jan 24, 2024. It is now read-only.
Add FLAGS_cinn_use_new_fusion_pass to switch to the new op fusion and fusion merge pass. #751
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
|
Thanks for your contribution! |
Xreki
force-pushed
the
switch_new_fusion_pass
branch
from
6001ac2 to
3fbd0aa
Compare
Xreki
force-pushed
the
switch_new_fusion_pass
branch
from
3edc77a to
7ead49b
Compare
wzzju
reviewed
Apr 25, 2022
wzzju
previously approved these changes
Apr 25, 2022
Xreki
changed the title
haozech
added a commit
that referenced
this pull request
Apr 27, 2022
* fix fusion pass (#737) * Fix Fusion Pass * fix auto-tune header export (#747) * Add the op mapper for fill_any_like. (#745) * Add the Optimize API. (#750) * Add the Optimize API. * Fix some pass applying errors in gemm_rewriter_test. * Use DefaultTrainingOptimizeOptions instead of DefaultOptimizeOptions. * Use cublas gemm instead of matmul. (#735) * Rewrite the left single matmul, use cublas call instead. * Simplify the code logic. * Add FLAGS_cinn_use_new_fusion_pass to switch to the new op fusion and fusion merge pass. (#751) * Add scatter_add op (#738) * add scatter_add base op * add some annotation * rename index_assign to scatter_assign (#739) * rename index_assign to scatter_assign * opmapper keep index_assign * add gather/scatter_add opmapper and change attr dtype to int64 (#743) * add gather/scatter_add opmapper and change attr dtype to int64 * Fix Fusion Pass And Lowering For BN And Laplace Model (#746) * Fix Fusion Pass For BN And Laplace Model * change arg arr of cuda kernel to vector<void*> (#757) * WIP switch ir schedule (#740) * update temp buffer and enable vectorize/unroll after ir schedule (#748) * Add new schedule and fix bugs (#752) * add new schedule * fix bugs add conv2d schedule (#760) Co-authored-by: sunli <sunli_hit@outlook.com> Co-authored-by: TeFeng Chen <ctfeng66@163.com> Co-authored-by: Zhen Wang <wangzhen31@baidu.com> Co-authored-by: Yiqun Liu <liuyiqun01@baidu.com> Co-authored-by: jiangcheng <thisjiang@qq.com> Co-authored-by: wangone <2279939962@qq.com>
zhhsplendid
pushed a commit
that referenced
this pull request
Jun 9, 2022
… fusion merge pass. (#751)
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
添加2个FLAGS，方便调试：
FLAGS_cinn_use_new_fusion_pass，切换到新的OpFusionPass+FusionMerge，默认值为false。FLAGS_cinn_sync_run，在每个Instruction->Run调用介绍后加入stream同步，用于debug，默认值为false。