titu1994 / keras-one-cycle

Implementation of One-Cycle Learning rate policy (adapted from Fast.ai lib)
MIT License
285 stars 78 forks source link

tensorflow 2.0 #13

Closed happiness-ai closed 5 years ago

happiness-ai commented 5 years ago

File "F:/workspace/python/tf-fit/keras_clr/clr.py", line 603, in inp = Input(shape=(10,)) File "E:\Anaconda3_5_0_0\envs\DeepLearning\lib\site-packages\keras\engine\input_layer.py", line 178, in Input input_tensor=tensor) File "E:\Anaconda3_5_0_0\envs\DeepLearning\lib\site-packages\keras\legacy\interfaces.py", line 91, in wrapper return func(*args, **kwargs) File "E:\Anaconda3_5_0_0\envs\DeepLearning\lib\site-packages\keras\engine\inputlayer.py", line 39, in init name = prefix + '' + str(K.get_uid(prefix)) File "E:\Anaconda3_5_0_0\envs\DeepLearning\lib\site-packages\keras\backend\tensorflow_backend.py", line 74, in get_uid graph = tf.get_default_graph() AttributeError: module 'tensorflow' has no attribute 'get_default_graph'

titu1994 commented 5 years ago

Keras does not support Tensorflow 2.0 as of version 2.2.4. You'll need to wait until TF 2.0 is officially released, and then Keras master is updated.