-
Notifications
You must be signed in to change notification settings - Fork 227
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
Fix exponential scheduler with bias #589
Fix exponential scheduler with bias #589
Conversation
Jenkins please retry a build |
@@ -194,11 +202,12 @@ def get_b(a, k): | |||
return p_min - a |
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.
def get_b(a, k): -> def get_b(a):
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.
Done.
@mkaglins @AlexKoff88 I suppose we could re-run some filter pruning benchmarks to see diff in accuracy for this scheduler once this fix is introduced? |
This |
return 1 / 3 * y + 1 / (y ** 7) - 4 / 3 | ||
c = (0.75 * p_max - p_min) / (p_max - p_min) | ||
y = np.exp(-x * epoch_idx) | ||
return y ** factor - c * y + c - 1 |
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.
Should we make the factor
parameter adjustable via config? I suppose the 1/8 value is based on what is set in the paper, not on our empirical results.
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.
Yes, it is a good idea to set better default value based on empirical results on our models (firstly get these results). But I am not sure about the importance of this param in the config for users.
* exponential_with_bias scheduler was fixed * Minor refactoring
Looks like that current implementation does not work properly.
Consider the following example:
I expected that
x
is equal to0.7
but I got0.666666666666667
.