Closed zaccharieramzi closed 2 years ago
Ack, just as reference, the original paper stated that
"The training was regularised by weight decay (the L2 penalty multiplier set to 5 · 10−4) and dropout regularisation for the first two fully-connected layers (dropout ratio set to 0.5)."
Feel free to send any PR if you would like to contribute. Thanks
Closing as per this.
If you open a GitHub issue, here is our policy:
It must be a bug, a feature request, or a significant problem with the documentation (for small docs fixes please send a PR instead). The form below must be filled out.
Here's why we have that policy:.
Keras developers respond to issues. We want to focus on work that benefits the whole community, e.g., fixing bugs and adding features. Support only helps individuals. GitHub also notifies thousands of people when issues are filed. We want them to see you communicating an interesting problem, rather than being redirected to Stack Overflow.
System information.
TensorFlow version (you are using): 2.8 Are you willing to contribute it (Yes/No) : Yes
Describe the feature and the current behavior/state.
I would like to be able to use dropout on the first two fully connected layers of the VGG implementations in keras applications, as specified in the original paper (and done in PyTorch).
Will this change the current api? How?
Yes: this will mean adding a
dropout
kwarg to the VGG implementations, that will default to0.0
to remain a non-breaking change.Who will benefit from this feature?
Users willing to reproduce the original results of the VGG paper.
Contributing