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
{{ message }}
This repository has been archived by the owner on Nov 17, 2023. It is now read-only.
I think it would be nice to make easier to get the test and validation error when using the Module API.
Right now you have to go through a convoluted process of using eval_end_callback and batch_end_callback and collapsing the batches so they match epochs. (Please correct me if there's an easier way to do this).
The goal would be to have a boolean toogle for the Module API which will record the train and validation error or have an specific callback that does this.
The text was updated successfully, but these errors were encountered:
I will send a patch with a new callback in the Module API to pass the eval metric for train and validation per epoch, any inconvenients with this approach?
I think it would be nice to make easier to get the test and validation error when using the Module API.
Right now you have to go through a convoluted process of using eval_end_callback and batch_end_callback and collapsing the batches so they match epochs. (Please correct me if there's an easier way to do this).
The goal would be to have a boolean toogle for the Module API which will record the train and validation error or have an specific callback that does this.
The text was updated successfully, but these errors were encountered: