In keras how to get epoch and validation loss from model checkpoint?
up vote
0
down vote
favorite
From here (https://keras.io/callbacks/#modelcheckpoint) you can save the best model according to the validation error by setting save_best_only.
I know you can save the corresponding epoch and validation error by writing it to the checkpoint's file name. However, this means a lot of models could be saved and I expect this to result in memory error on my GPU.
Is there a way to get the epoch and val loss corresponding to the final best model without having to write it in the filename?
keras
asked Nov 19 at 18:12
user5490
84
add a comment |
up vote
0
down vote
favorite
From here (https://keras.io/callbacks/#modelcheckpoint) you can save the best model according to the validation error by setting save_best_only.
I know you can save the corresponding epoch and validation error by writing it to the checkpoint's file name. However, this means a lot of models could be saved and I expect this to result in memory error on my GPU.
Is there a way to get the epoch and val loss corresponding to the final best model without having to write it in the filename?
keras
asked Nov 19 at 18:12
user5490
84
add a comment |
up vote
0
down vote
favorite
up vote
0
down vote
favorite
From here (https://keras.io/callbacks/#modelcheckpoint) you can save the best model according to the validation error by setting save_best_only.
I know you can save the corresponding epoch and validation error by writing it to the checkpoint's file name. However, this means a lot of models could be saved and I expect this to result in memory error on my GPU.
Is there a way to get the epoch and val loss corresponding to the final best model without having to write it in the filename?
keras
asked Nov 19 at 18:12
user5490
84
From here (https://keras.io/callbacks/#modelcheckpoint) you can save the best model according to the validation error by setting save_best_only.
I know you can save the corresponding epoch and validation error by writing it to the checkpoint's file name. However, this means a lot of models could be saved and I expect this to result in memory error on my GPU.
Is there a way to get the epoch and val loss corresponding to the final best model without having to write it in the filename?
keras
keras
asked Nov 19 at 18:12
user5490
84
asked Nov 19 at 18:12
user5490
84
asked Nov 19 at 18:12
user5490
84
asked Nov 19 at 18:12
user5490
84
asked Nov 19 at 18:12
user5490
84
84
add a comment |
add a comment |
1 Answer
1
active
oldest
votes
up vote
0
down vote
Depends on what you want to do with the epoch and validation error and at what time in the training, but you can implement your own callback functionality quite easily. What you want to get the training metrics is the logs object, which is passed to the callback in each of the callback events (see here).
If for example you need to call a certain function f with the epoch and validation error at the end of every epoch, you could implement this using the LambdaCallback:
keras.callbacks.LambdaCallback(on_epoch_end=lambda epoch, logs: f(epoch, logs['val_loss']))
If instead you want to use the ModelCheckpoint callback but don't want it to write to file, you can create a custom CallBack that rewrites the ModelCheck callback and changes the save behavior (code here).
(Don't know if that answers your question, not entirely sure what the requirements are)
answered Nov 20 at 0:04
lmartens
647517
add a comment |
1 Answer
1
active
oldest
votes
1 Answer
1
active
oldest
votes
active
oldest
votes
active
oldest
votes
up vote
0
down vote
Depends on what you want to do with the epoch and validation error and at what time in the training, but you can implement your own callback functionality quite easily. What you want to get the training metrics is the logs object, which is passed to the callback in each of the callback events (see here).
If for example you need to call a certain function f with the epoch and validation error at the end of every epoch, you could implement this using the LambdaCallback:
keras.callbacks.LambdaCallback(on_epoch_end=lambda epoch, logs: f(epoch, logs['val_loss']))
If instead you want to use the ModelCheckpoint callback but don't want it to write to file, you can create a custom CallBack that rewrites the ModelCheck callback and changes the save behavior (code here).
(Don't know if that answers your question, not entirely sure what the requirements are)
answered Nov 20 at 0:04
lmartens
647517
add a comment |
up vote
0
down vote
Depends on what you want to do with the epoch and validation error and at what time in the training, but you can implement your own callback functionality quite easily. What you want to get the training metrics is the logs object, which is passed to the callback in each of the callback events (see here).
If for example you need to call a certain function f with the epoch and validation error at the end of every epoch, you could implement this using the LambdaCallback:
keras.callbacks.LambdaCallback(on_epoch_end=lambda epoch, logs: f(epoch, logs['val_loss']))
If instead you want to use the ModelCheckpoint callback but don't want it to write to file, you can create a custom CallBack that rewrites the ModelCheck callback and changes the save behavior (code here).
(Don't know if that answers your question, not entirely sure what the requirements are)
answered Nov 20 at 0:04
lmartens
647517
add a comment |
up vote
0
down vote
up vote
0
down vote
Depends on what you want to do with the epoch and validation error and at what time in the training, but you can implement your own callback functionality quite easily. What you want to get the training metrics is the logs object, which is passed to the callback in each of the callback events (see here).
If for example you need to call a certain function f with the epoch and validation error at the end of every epoch, you could implement this using the LambdaCallback:
keras.callbacks.LambdaCallback(on_epoch_end=lambda epoch, logs: f(epoch, logs['val_loss']))
If instead you want to use the ModelCheckpoint callback but don't want it to write to file, you can create a custom CallBack that rewrites the ModelCheck callback and changes the save behavior (code here).
(Don't know if that answers your question, not entirely sure what the requirements are)
answered Nov 20 at 0:04
lmartens
647517
Depends on what you want to do with the epoch and validation error and at what time in the training, but you can implement your own callback functionality quite easily. What you want to get the training metrics is the logs object, which is passed to the callback in each of the callback events (see here).
If for example you need to call a certain function f with the epoch and validation error at the end of every epoch, you could implement this using the LambdaCallback:
keras.callbacks.LambdaCallback(on_epoch_end=lambda epoch, logs: f(epoch, logs['val_loss']))
If instead you want to use the ModelCheckpoint callback but don't want it to write to file, you can create a custom CallBack that rewrites the ModelCheck callback and changes the save behavior (code here).
(Don't know if that answers your question, not entirely sure what the requirements are)
answered Nov 20 at 0:04
lmartens
647517
answered Nov 20 at 0:04
lmartens
647517
answered Nov 20 at 0:04
lmartens
647517
answered Nov 20 at 0:04
lmartens
647517
647517
add a comment |
add a comment |
Thanks for contributing an answer to Stack Overflow!
- Please be sure to answer the question. Provide details and share your research!
But avoid …
- Asking for help, clarification, or responding to other answers.
- Making statements based on opinion; back them up with references or personal experience.
To learn more, see our tips on writing great answers.
Some of your past answers have not been well-received, and you're in danger of being blocked from answering.
Please pay close attention to the following guidance:
- Please be sure to answer the question. Provide details and share your research!
But avoid …
- Asking for help, clarification, or responding to other answers.
- Making statements based on opinion; back them up with references or personal experience.
To learn more, see our tips on writing great answers.
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
StackExchange.ready(
function () {
StackExchange.openid.initPostLogin('.new-post-login', 'https%3a%2f%2fstackoverflow.com%2fquestions%2f53380417%2fin-keras-how-to-get-epoch-and-validation-loss-from-model-checkpoint%23new-answer', 'question_page');
}
);
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Sign up or log in
StackExchange.ready(function () {
StackExchange.helpers.onClickDraftSave('#login-link');
});
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Sign up using Google
Sign up using Facebook
Sign up using Email and Password
Post as a guest
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
Required, but never shown
