This repository has been archived by the owner on Nov 17, 2023. It is now read-only.
[contrib][op] fix MultiBoxPrior confusing results if first ratio is not 1.0 #13763
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
|
@mxnet-label-bot Add [Operator, pr-awaiting-review] |
zhreshold
force-pushed
the
fix-multibox-prior-logic
branch
from
408f7a1
to
9220e31
Compare
|
@eric-haibin-lin could you review it? Thanks |
|
@zhreshold - Thanks. |
zhreshold
force-pushed
the
fix-multibox-prior-logic
branch
from
9220e31
to
871d250
Compare
vandanavk
reviewed
Feb 5, 2019
| @@ -297,6 +297,18 @@ def f(x, a, b, c): | |||
| [backward_expected], | |||
| rtol=1e-2 if dtype is np.float16 else 1e-5, | |||
| atol=1e-2 if dtype is np.float16 else 1e-5) | |||
| def test_multibox_prior_op(): | |||
| h = 561 | |||
| w = 728 | |||
There was a problem hiding this comment.
Is it possible to not hardcode the test values?
There was a problem hiding this comment.
the input shape is designed specifically for the network
|
@eric-haibin-lin ping for review! |
There was a problem hiding this comment.
@zhreshold can you please explain why we need the changes in the onnx-tensorrt submodule in this PR?
zhreshold
force-pushed
the
fix-multibox-prior-logic
branch
from
871d250
to
a7de725
Compare
|
Should be fine, I've tried rebasing to master |
|
@zhreshold Could you please provide an update on this PR? It seems no tracking for last 2 weeks. Thanks ! |
zhreshold
force-pushed
the
fix-multibox-prior-logic
branch
2 times, most recently
from
edee717
to
be55418
Compare
zhreshold
force-pushed
the
fix-multibox-prior-logic
branch
from
be55418
to
17aba5e
Compare
zhreshold
force-pushed
the
fix-multibox-prior-logic
branch
from
17aba5e
to
e4c6c16
Compare
kedarbellare
pushed a commit
to kedarbellare/incubator-mxnet
that referenced
this pull request
Apr 20, 2019
…ot 1.0 (apache#13763) * fix confusion results if ratios are set differently * update 3rdparty
haohuanw
pushed a commit
to haohuanw/incubator-mxnet
that referenced
this pull request
Jun 23, 2019
…ot 1.0 (apache#13763) * fix confusion results if ratios are set differently * update 3rdparty
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.
Description
Replace a hardcoded ratio in MultiBoxPrior with whatever user desired.
It was hardcoded to 1 by default, now it allow arbitrary value, even though ratio==1.0 is highly recommended in almost all circumstances for computer vision tasks.
This change does NOT affecting existing examples and use cases.
Checklist
Essentials
Please feel free to remove inapplicable items for your PR.