-
Notifications
You must be signed in to change notification settings - Fork 409
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
Softsign optimization #585
Softsign optimization #585
Conversation
Looks good to me. Would be nice to add a test, just to show correct behaviour. We could do it in a similar way to other activations, in |
Added one test file. |
test/pytest/test_softsign.py
Outdated
|
||
print('Accuracy hls4ml relative to keras: {}'.format(acc_hls4ml)) | ||
|
||
assert acc_hls4ml >= 0.7 |
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.
Why only 0.7?
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.
Changed to 0.98. New tests have also been added to the pytest file.
Quartus Streaming Softsign (PR #585 contd.)
* Softsign LUT optimization * Test file: Activation softsign * Changing minimal accurancy to 9.8 and adding new texts using Vivado and Quartus Co-authored-by: Nemer Chiedde <[email protected]>
Quartus Streaming Softsign (PR fastmachinelearning#585 contd.)
Storing only the positive part of the Softsign is possible due to its symmetry in the Y axis. Same theory when presented in pull request: #527