-
Notifications
You must be signed in to change notification settings - Fork 30
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
Added GoogleNet (Inception V1) Model #47
Conversation
@juliagsy I have implemented the model, it's test function, and I have also applied the General Pipeline guide, please review it and let me know if there are issues in it. |
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.
Hey Sarvesh! Thanks for the great work! I've left some comments below mainly related to support general additional support and refactor. The updated guidelines can also be found here: https://docs.google.com/document/d/1NfVngAQlQQlxLGEg47O2lnfn60M5xZekZFei9k5RJd8/edit?usp=sharing
Please feel free to lmk if you have any questions! Thanks!
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.
Hey Sarvesh! Thanks for the great work! The code looks good to me overall and we're good to merge once the final comments are addressed and that the tests are passing! Thanks!
PS: please use "squash and merge" when merging :D
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.
Hey Sarvesh! All dropout values should be 0 including aux_dropout. I've left some comments, once they are all in place, the test should pass for both load weights = True and False (I've tried it out :D) Thanks!
PS: Please also clean up the code after that and we'll then be able to look into merging, thanks!
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.
Hey Sarvesh! Thanks for the changes! The PR may not be ready yet, but just thought I'd leave a comment before I forget, please remove any changes unrelated to models and keep googlenet related updates only, thanks!
31e69ed
to
e8be130
Compare
Sure will take care from next time. |
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.
Hey Sarvesh! I was about to merge it and doing a final check, but the pretrained weights loading result is not correct anymore. I'm not sure what changed (most likely the architecture and its config) but could you please double check (one thing is that I had to make fc
with_bias=True for the weights mapping to map correctly but I didn't have to change anything previously to make the tests run)
I'm also pretty sure that during the last review, tests were passing with minimal change (the review comments), let me know once it's done, thanks!
Hey! A quick update, seems like inception3 had a similar issue 🤔 |
Working on the resnet failure issue. |
e2938db
to
fe214a3
Compare
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.
I'm temp commenting out logins value tests (the float dtype ones)
otherwise lgtm, thanks!
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.
lgtm, thanks!
Corresponding to: #38