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

Draft: [onert-micro] Bring up SparseCrossEntropy Loss #13431

Draft
wants to merge 5 commits into
base: master
Choose a base branch
from
Draft

Draft: [onert-micro] Bring up SparseCrossEntropy Loss #13431

wants to merge 5 commits into from

Conversation

ljwoo94
Copy link
Contributor

@ljwoo94 ljwoo94 commented Jul 15, 2024 •
edited
Loading

What

  • This commit adds feature to support Sparse Cross Entropy as a Loss function.
  • This commit also adds Sparse Cross Entropy Accuracy metric to check accuracy of above loss.

ToDo

Jungwoo Lee added 4 commits July 11, 2024 15:58
DRAFT: [onert-micro] Add SimpleMnistClassificationTask Test
ac1f3a0 
This testcase is added to test SparseCrossEntropy featture that will be
added

Signed-off-by: Jungwoo Lee <[email protected]>
Add Test Case to run
d45726c
Test Draft with mnist Data
9c63433
Add Sparse Cross Entropy
a52d789
@ljwoo94
Copy link
Contributor Author

ljwoo94 commented Jul 15, 2024

TF Result: .95
Onert-micro result: .95

BalyshevArtem
BalyshevArtem approved these changes Jul 19, 2024
View reviewed changes
Copy link
Contributor

@BalyshevArtem BalyshevArtem left a comment

Choose a reason for hiding this comment

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

overall lgtm, thank you!
Let's split this draft into prs and merge it

ljwoo94
ljwoo94 commented Jul 22, 2024
View reviewed changes
onert-micro/onert-micro/src/core/train/OMTrainingHandler.cpp
Comment on lines +67 to +70
if (loss_type == SPARSE_CROSS_ENTROPY)
{
offset = batch_num * data_type_size;
}
Copy link
Contributor Author

Choose a reason for hiding this comment

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

@BalyshevArtem
Is it ok to changing offset according to the loss_type??
I had no idea of refactoring this code to apply sparse cross entropy... :'(

@ljwoo94 ljwoo94 mentioned this pull request Aug 9, 2024
Open
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Reviewers

@BalyshevArtem BalyshevArtem BalyshevArtem approved these changes

@chunseoklee chunseoklee Awaiting requested review from chunseoklee

Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

Successfully merging this pull request may close these issues.
2 participants
@ljwoo94 @BalyshevArtem