wassname/ray
1
0
Fork 0
mirror of https://github.com/wassname/ray.git synced 2026-06-27 20:38:19 +08:00
Code Issues Packages Projects Releases Wiki Activity

[tune] Fix bug in example where config hyperparameters were ignored (#2860)

Browse Source 
A fix to an example for tune (`python/ray/tune/examples/pbt_tune_cifar10_with_keras.py`) where the hyperparameters for the optimizer, learning rate and decay, were not being passed into the optimizer. 

This means that the current optimizer uses default values for the hyperparameters no matter the config.
This commit is contained in:
Daniel Ho
2018-09-12 09:17:56 -07:00
committed by Richard Liaw
parent f3c1194be3
commit d9eeaaf00a
1 changed files with 2 additions and 1 deletions
Download Patch File Download Diff File Expand all files Collapse all files
python/ray/tune/examples/pbt_tune_cifar10_with_keras.py
+2 -1
View File
@@ -110,7 +110,8 @@ class Cifar10Model(Trainable):
x_train = self.train_data[0]
model = self._build_model(x_train.shape[1:])
opt = tf.keras.optimizers.Adadelta()
opt = tf.keras.optimizers.Adadelta(
lr=self.config["lr"], decay=self.config["decay"])
model.compile(
loss="categorical_crossentropy",
optimizer=opt,