Timeline for Keras difference beetween val_loss and loss during training

13 events

when toggle format	what		by	license	comment
Mar 17, 2023 at 19:04	comment	added	Green Falcon		You can, but the easier solution is to add them when you want to create the layers of your model. You can use `kernel_regularizer` and `bias_regularizer`. You can access l1 through `tf.keras.regularizers.l1`
Mar 17, 2023 at 11:48	comment	added	Antonios Sarikas		@GreenFalcon Should the validation loss include the penalization (e.g. $\ell_1$ norm) we impose on the training loss?
Jun 16, 2020 at 9:43	comment	added	Green Falcon		Would you elaborate?
Jun 16, 2020 at 0:51	comment	added	Kermit		This is misleading because it does not have to be about `cross-validation`
S Jan 21, 2020 at 13:17	history	suggested	sophros	CC BY-SA 4.0	Code and quote formatting
Jan 21, 2020 at 11:16	review	Suggested edits
S Jan 21, 2020 at 13:17
May 1, 2019 at 13:59	history	edited	Green Falcon	CC BY-SA 4.0	added 5 characters in body
Dec 31, 2017 at 13:22	history	edited	Green Falcon	CC BY-SA 3.0	deleted 64 characters in body
Dec 31, 2017 at 13:01	history	edited	Green Falcon	CC BY-SA 3.0	added 1631 characters in body; deleted 1 character in body
Dec 31, 2017 at 12:56	history	edited	Green Falcon	CC BY-SA 3.0	added 1631 characters in body; deleted 1 character in body
Dec 31, 2017 at 12:40	history	edited	Green Falcon	CC BY-SA 3.0
Nov 30, 2017 at 20:23	review	Low quality posts
Nov 30, 2017 at 20:26
Nov 30, 2017 at 20:03	history	answered	Green Falcon	CC BY-SA 3.0