APTx activation function #214
Merged
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
docs/advanced.ipynb
Outdated
There was a problem hiding this comment.
Deleted by mistake?
There was a problem hiding this comment.
wow, i guess so. I thought only added 2 lines. it's been a while since i used git. i must have made a mistake.
There was a problem hiding this comment.
hi, i removed the modifications to the markdown in advanced.ipynb since vs code automatically delete those 3k lines for no reasons and i don't know why at the moment.
Rockdeldiablo
force-pushed
the
APTx_act
branch
from
ce337d3
to
bf4210d
Compare
shuheng-liu
approved these changes
May 14, 2024
| self.gamma = gamma | ||
|
|
||
| def forward(self, x): | ||
| return (self.alpha + torch.nn.functional.tanh(self.beta*x))*self.gamma*x |
There was a problem hiding this comment.
Optionally, you can just use torch.tanh(). If I recall correctly, torch.nn.functional is deprecated. I'll merge this one though.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
I saw you implemented the Swish function. As an exercise i implemented the APTx activation https://arxiv.org/abs/2209.06119 , which seems to be a variant of the new mish function that requires less computational power. I tested on the DE given as exercise during the Harvard ComputeFest and it seems to give better results than Swish at low/very low epochs. However more tests are needed before a definitive response.