-
Notifications
You must be signed in to change notification settings - Fork 88
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
Dont insert reshapes when converting pooling to reduce #2149
Conversation
Codecov Report
@@ Coverage Diff @@
## develop #2149 +/- ##
========================================
Coverage 91.53% 91.53%
========================================
Files 429 429
Lines 16011 16009 -2
========================================
- Hits 14655 14654 -1
+ Misses 1356 1355 -1
|
Performance tests have compilation errors, can you look into it ? |
This build is not recommended to merge 🔴 |
🔴torchvision-inceptionv3_1: FAILED: MIGraphX is not within tolerance - check verbose output🔴cadene-dpn92_1: FAILED: MIGraphX is not within tolerance - check verbose output🔴slim-inceptionv4_1: FAILED: MIGraphX is not within tolerance - check verbose output |
@pfultz2 accuracy tests are still failing during compile. Can you check ? |
@umangyadav I pushed a fix. |
instruction_ref pooling{}; | ||
|
||
std::vector<std::int64_t> axes(lens.size() - 2); | ||
std::iota(axes.begin(), axes.end(), 2); |
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.
Should this be doing any checks for the Layout ?
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.
@pfultz2 sorry i missed this comment before approving. Would this require check on the layout ?
@@ -122,6 +122,11 @@ struct find_nop_reshapes | |||
reshapes.insert("pad"); | |||
reshapes.insert("slice"); | |||
reshapes.insert("transpose"); | |||
reshapes.insert("reduce_mean"); |
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.
Would be better to assert that axes are either empty resulting in no-op or
axes lengths are 1.
is it improving reduce fusion ? Which of the matcher is helping doing that ? |
This should improve fusions.