tf.keras.callbacks. Reduce LROn Plateau
GPTKB entity
Statements (41)
| Predicate | Object |
|---|---|
| gptkbp:instance_of |
gptkb:API
|
| gptkbp:affects |
Helps to avoid overfitting by reducing the learning rate.
Can lead to better model performance. Improves model convergence. |
| gptkbp:description |
A Keras callback that reduces the learning rate when a metric has stopped improving.
|
| https://www.w3.org/2000/01/rdf-schema#label |
tf.keras.callbacks. Reduce LROn Plateau
|
| gptkbp:inherits_from |
gptkb:API
|
| gptkbp:introduced_in |
gptkb:Tensor_Flow_2.0
|
| gptkbp:is_designed_to |
cooldown=0
factor=0.1 min_delta=0.0001 min_lr=0. mode='auto' monitor='val_loss' patience=10 verbose=0 |
| gptkbp:orbital_period |
gptkb:monitor
gptkb:video_game patience verbose factor cooldown min_delta min_lr |
| gptkbp:related_to |
gptkb:Keras
gptkb:neural_networks gptkb:Deep_Learning Model Training Learning Rate Scheduling |
| gptkbp:technique |
reset
on_epoch_end on_train_end get_lr get_monitor_value on_batch_end on_train_begin set_model |
| gptkbp:type_of |
Callback Class
|
| gptkbp:usage |
Used to adjust the learning rate during training.
|
| gptkbp:bfsParent |
gptkb:Tensor_Flow
|
| gptkbp:bfsLayer |
4
|