Skip to content

fix bias size assert#5992

Merged
apaszke merged 5 commits intopytorch:masterfrom
elanmart:bilinear-fix
Mar 27, 2018
Merged

fix bias size assert#5992
apaszke merged 5 commits intopytorch:masterfrom
elanmart:bilinear-fix

Conversation

@elanmart
Copy link
Copy Markdown
Contributor

Fixes #5990

@elanmart
Copy link
Copy Markdown
Contributor Author

@pytorchbot retest this please

Comment thread aten/src/ATen/native/Linear.cpp Outdated

This comment was marked as off-topic.

This comment was marked as off-topic.

@apaszke
Copy link
Copy Markdown
Contributor

apaszke commented Mar 25, 2018

Also, can you please add a test for Bilinear without bias? Just to make sure we won't have a regression again

ezyang
ezyang previously requested changes Mar 26, 2018
Copy link
Copy Markdown
Contributor

@ezyang ezyang left a comment

Choose a reason for hiding this comment

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

See @apaszke

@elanmart
Copy link
Copy Markdown
Contributor Author

@ezyang @apaszke Please excuse me the delay.

How should I read the onnx CI failure?

@soumith
Copy link
Copy Markdown
Collaborator

soumith commented Mar 26, 2018

@elanmart the onnx failure means it's asking you to rebase :)

@elanmart
Copy link
Copy Markdown
Contributor Author

@pytorchbot retest this please.

@elanmart
Copy link
Copy Markdown
Contributor Author

The onnx failure seems unrelated [1], but I can't force the tests to run again :(

[1]

E: Failed to fetch http://dl.bintray.com/apache/cassandra/dists/39x/main/binary-amd64/Packages  502  Bad Gateway [IP: 52.27.175.225 80]

@apaszke
Copy link
Copy Markdown
Contributor

apaszke commented Mar 27, 2018

@onnxbot retest this please

Comment thread test/test_nn.py
_assertGradAndGradgradChecks(self, lambda x1, x2: F.bilinear(x1, x2, module.weight, module.bias),
(input1_1, input2_1))

def test_bilinear_no_bias(self):

This comment was marked as off-topic.

This comment was marked as off-topic.

@ezyang ezyang dismissed their stale review March 27, 2018 14:51

new commit

@ezyang
Copy link
Copy Markdown
Contributor

ezyang commented Mar 27, 2018

@pytorchbot retest this please

Comment thread test/test_nn.py Outdated
output_no_bias = module_no_bias(input1, input2)

self.assertEqual(output, output_no_bias)
output_no_bias.sum().backward()

This comment was marked as off-topic.

@apaszke apaszke merged commit 5583c12 into pytorch:master Mar 27, 2018
@elanmart elanmart deleted the bilinear-fix branch March 27, 2018 22:17
laurentdupin pushed a commit to laurentdupin/pytorch that referenced this pull request Apr 24, 2026
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Projects

None yet

Development

Successfully merging this pull request may close these issues.

nn.Bilinear without bias broken

4 participants