[CINN] Add new group scheduler #56444
Merged
Conversation
This file contains 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
|
你的PR提交成功，感谢你对开源项目的贡献! |
|
Sorry to inform you that 1532568's CIs have passed for more than 7 days. To prevent PR conflicts, you need to re-run all CIs manually. |
|
Sorry to inform you that 0763775's CIs have passed for more than 7 days. To prevent PR conflicts, you need to re-run all CIs manually. |
|
Sorry to inform you that 07f5674's CIs have passed for more than 7 days. To prevent PR conflicts, you need to re-run all CIs manually. |
zhhsplendid
reviewed
Nov 1, 2023
paddle/cinn/auto_schedule/search_space/auto_gen_rule/reduction_factoring.cc
Show resolved
Hide resolved
zeroRains
pushed a commit
to zeroRains/Paddle
that referenced
this pull request
Nov 8, 2023
* [CINN] Add new group scheduler * [Fix] Fix priority of bind loops and output node * [Fix] Set is_reduce_axis for Argmin and Argmax * [Fix] Add producer consumer relation to Argmax * [Fix] Add NodePriority and skip ExternCall block * [Fix] Add prohibit schedule block * [Fix] schedule block graph test * [Fix] Skip external calls while auto inline * [Fix] Fix relationship of block with Call nodes * [CINN] Use common reduce while loop fusion * [Fix] ScheduleBlockGraph unittest * [Fix] reduction unittests * [Fix] Skip group schedule of NonFusible nodes * [Fix] Incomplete AutoSimplify * [Fix] Adapt to new GraphCompilers * [Fix] schedule block graph unittest * [Fix] loop reorder to match master * [Fix] elementwise loop alignment * [Fix] cuda axis coeff and range * [Fix] Add conditions to schedules related to cuda * [Fix] fix conflict * fix conflict * [Fix] fix conflict * Integrate ReductionFactoring * [CINN] Upgrade ReductionReduction rule * resolve conflict * fix tensor in wb-block * add reduce type in FactorizeReduction * [CINN] Add cross thread reduction replacer * Integrate cross thread reduction * add anonymous namespace * fix reduction factoring unittest * Prohibit group schedule on single op * Revert "Prohibit group schedule on single op" This reverts commit 13ddff9. * fix reduction factoring unittest * fix reduction factoring unittest * open group scheduler flag * fix node priority * fix cross thread reduction on cpu * fix reduction_factoring with pre Fuse * Revert "open group scheduler flag" This reverts commit 192ccc1. * Revert "fix reduction_factoring with pre Fuse" This reverts commit 31889eb. * simplify log of range * add a TODO * fix x86 reduction bug
danleifeng
pushed a commit
to danleifeng/Paddle
that referenced
this pull request
Nov 14, 2023
* [CINN] Add new group scheduler * [Fix] Fix priority of bind loops and output node * [Fix] Set is_reduce_axis for Argmin and Argmax * [Fix] Add producer consumer relation to Argmax * [Fix] Add NodePriority and skip ExternCall block * [Fix] Add prohibit schedule block * [Fix] schedule block graph test * [Fix] Skip external calls while auto inline * [Fix] Fix relationship of block with Call nodes * [CINN] Use common reduce while loop fusion * [Fix] ScheduleBlockGraph unittest * [Fix] reduction unittests * [Fix] Skip group schedule of NonFusible nodes * [Fix] Incomplete AutoSimplify * [Fix] Adapt to new GraphCompilers * [Fix] schedule block graph unittest * [Fix] loop reorder to match master * [Fix] elementwise loop alignment * [Fix] cuda axis coeff and range * [Fix] Add conditions to schedules related to cuda * [Fix] fix conflict * fix conflict * [Fix] fix conflict * Integrate ReductionFactoring * [CINN] Upgrade ReductionReduction rule * resolve conflict * fix tensor in wb-block * add reduce type in FactorizeReduction * [CINN] Add cross thread reduction replacer * Integrate cross thread reduction * add anonymous namespace * fix reduction factoring unittest * Prohibit group schedule on single op * Revert "Prohibit group schedule on single op" This reverts commit 13ddff9. * fix reduction factoring unittest * fix reduction factoring unittest * open group scheduler flag * fix node priority * fix cross thread reduction on cpu * fix reduction_factoring with pre Fuse * Revert "open group scheduler flag" This reverts commit 192ccc1. * Revert "fix reduction_factoring with pre Fuse" This reverts commit 31889eb. * simplify log of range * add a TODO * fix x86 reduction bug
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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.
PR types
New features
PR changes
Others
Description
card-74457
Add GroupScheduler to schedule fusion groups.
Its responsibility is to perform loop alignment, automatic inline, automatic loop fusion, and optimize the storage location of intermediate variables.