You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I observed the strange behavior during the fine-tuning of MobileNet v2 when doing the filter pruning. I set the "pruning_target" to 0.3 and "pruning_steps" to15 but I didn't get the desired ratio after 15 epochs. It was achieved after the 25th epoch.
The problem affects only exponential and exponential_with_bias schedulers.
The problem was caused by the momentum parameter in the optimizer. Statistics of the momentum from earlier epochs (with smaller pruning rates) summed with weights on the next epochs (with higher pruning rates) and make them non-zero.
After a couple of training steps, this effect is vanishing and the pruning rate of masks becomes equal to the pruning rate of weights.
As a solution, it was decided to add mask applying on every training step (that significantly speeds up vanishing of momentum non-zero elements). Also, a warning message about this effect will be added to the release notes.
Changes are done in #365
I observed the strange behavior during the fine-tuning of MobileNet v2 when doing the filter pruning. I set the "pruning_target" to 0.3 and "pruning_steps" to15 but I didn't get the desired ratio after 15 epochs. It was achieved after the 25th epoch.
Please consider with high priority.
Here is the config:
The text was updated successfully, but these errors were encountered: