experiments/STT
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Add force_initialize_learning_rate

Browse Source 
Ability to reset learning rate which has been reduced by reduce_lr_on_plateau
This commit is contained in:
Richard Hamnett 2020-02-21 18:32:03 +00:00 committed by GitHub
parent aff310d73a
commit 0de9e4bf80
No known key found for this signature in database
GPG Key ID: 4AEE18F83AFDEB23
1 changed files with 1 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

1
util/flags.py
View File

@ -148,6 +148,7 @@ def create_flags():
f.DEFINE_boolean('reduce_lr_on_plateau', False, 'Enable reducing the learning rate if a plateau is reached. This is the case if the validation loss did not improve for some epochs.')
f.DEFINE_integer('plateau_epochs', 10, 'Number of epochs to consider for RLROP. Has to be smaller than es_epochs from early stopping')
f.DEFINE_float('plateau_reduction', 0.1, 'Multiplicative factor to apply to the current learning rate if a plateau has occurred.')
f.DEFINE_float('force_initialize_learning_rate', False, 'Force re-initialization of learning rate which was previously reduced.')
# Decoder