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
Hello everybody, I am using keras tuner version 1.0.1 with tensorflow 2.0.0 to fine tune the learning rate, batch-size and the number of epoch.
as you can see only one trial with learning rate=0.01, batch size=8 and number of epochs=80 took 7 hours until only the 32th and the colab bans me from using the gpu ( the google account is brand new, only used it for this notebook) and the results are very poor.
while without keras tuner 100 epochs doesn't even take 2 hours for this batch size and this learning rate and the training iou already get the value of 0.78 and val_iou=0.3 from the first epoch while for the first epoch with keras tuner :
Can someone help me with this problem?
N.B I used the subclass provided here in the documentation #122
The text was updated successfully, but these errors were encountered:
luna9722
changed the title
one trial take 7 hours and block in the 32th epoch
one trial took 7 hours and block in the 32th epoch
May 3, 2021
Hello everybody, I am using keras tuner version 1.0.1 with tensorflow 2.0.0 to fine tune the learning rate, batch-size and the number of epoch.
as you can see only one trial with learning rate=0.01, batch size=8 and number of epochs=80 took 7 hours until only the 32th and the colab bans me from using the gpu ( the google account is brand new, only used it for this notebook) and the results are very poor.
while without keras tuner 100 epochs doesn't even take 2 hours for this batch size and this learning rate and the training iou already get the value of 0.78 and val_iou=0.3 from the first epoch while for the first epoch with keras tuner :
Can someone help me with this problem?
N.B I used the subclass provided here in the documentation #122
The text was updated successfully, but these errors were encountered: