[BUG] Initializing any DL estimator with non-default values constructs these values only superficially #4067
Labels
bug
Something isn't working
module:classification
classification module: time series classification
module:regression
regression module: time series regression
Describe the bug
A logical bug is present in all DL estimators and is caused by not passing custom parameters to its underlying
Network
sinceNetwork
actually makes use of these values to create the DL model of appropriate size. Take a look at the example provided below.To Reproduce
Output:
Expected behavior
This is happening because of
self._network = CNNNetwork()
(or corresponding Network) in the constructor of the estimator. The values that we receive are to be passed into the Network.Additional context
I'm speculating this is the reason why the checks at #4036 are failing despite having very small hyperparameters. In any case, I'll open a PR to resolve this bug.
Versions
The text was updated successfully, but these errors were encountered: