Skip to content
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.

Sign up for GitHub

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

Jump to bottom

Define shape of TF lable tensor based on image tensor shape #142

Merged
merged 1 commit into from
Sep 14, 2018
Merged

Define shape of TF lable tensor based on image tensor shape #142

merged 1 commit into from
Sep 14, 2018

Conversation

JanuszL
Copy link
Contributor

@JanuszL JanuszL commented Aug 30, 2018

Signed-off-by: Janusz Lisiecki jlisiecki@nvidia.com

@JanuszL
Copy link
Contributor Author

JanuszL commented Aug 30, 2018

I think this path make valid assumptions - that labels dims in TF plugin are (batch_size, 1).
This helps with #140 where prefetch_queue from TF slim expects all tensors to have its shape defined. Now it is done only for image tensor.

@JanuszL JanuszL mentioned this pull request Aug 30, 2018
Closed
@JanuszL JanuszL requested review from pribalta and removed request for seanprime7 August 30, 2018 16:05
@JanuszL
Define shape of TF lable tensor based on image tensor shape
99244a1 
Signed-off-by: Janusz Lisiecki <jlisiecki@nvidia.com>
pribalta
pribalta approved these changes Sep 14, 2018
View reviewed changes
Copy link
Contributor

@pribalta pribalta left a comment

Choose a reason for hiding this comment

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

looks good to me

@JanuszL JanuszL merged commit 3ab7a13 into NVIDIA:master Sep 14, 2018
@JanuszL JanuszL deleted the TF_lable_shape branch September 14, 2018 08:44
@JanuszL JanuszL modified the milestones: Release_0.3.0, Release_0.4.0 Sep 20, 2018
JanuszL added a commit that referenced this pull request Oct 8, 2018
@JanuszL
Revert #142. It seems that TF doesn't like having lable tensor size a…
c8ff264 
…lways defined. (#203)

Somehow TF during training instead of having (n,) tensor redefines it as a (n, 1)
and complains that it cannot be 2D.

Signed-off-by: Janusz Lisiecki <jlisiecki@nvidia.com>
JanuszL added a commit that referenced this pull request Oct 8, 2018
@JanuszL
Revert #142. It seems that TF doesn't like having lable tensor size a…
b54b68b 
…lways defined. (#203)

Somehow TF during training instead of having (n,) tensor redefines it as a (n, 1)
and complains that it cannot be 2D.

Signed-off-by: Janusz Lisiecki <jlisiecki@nvidia.com>
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@ptrendx ptrendx ptrendx approved these changes

@pribalta pribalta pribalta approved these changes

Assignees

@ptrendx ptrendx

Labels
None yet
Projects
None yet
Milestone
Release_0.4.0
Development

Successfully merging this pull request may close these issues.

None yet
3 participants
@JanuszL @ptrendx @pribalta