Closed PonteIneptique closed 6 years ago
Thanks! What error do you get with tensorflow 1.7.0 ?
Here was the error
WARNING:tensorflow:From /usr/local/lib/python3.5/dist-packages/tensorflow/contrib/
learn/python/learn/datasets/base.py:198: retry
(from tensorflow.contrib.learn.python.learn.datasets.base) is deprecated and will
be removed in a future version.
Instructions for updating:
Use the retry module or similar alternatives.
WARNING - CRNN_experiment - No observers have been added to this run
INFO - CRNN_experiment - Running command 'run'
INFO - CRNN_experiment - Started
ERROR - CRNN_experiment - Failed after 0:00:00!
Traceback (most recent calls WITHOUT Sacred internals):
File "train.py", line 95, in run
train_distribute=distribution_gpus(parameters.num_gpus))
File "train.py", line 25, in distribution_gpus
return tf.contrib.distribute.OneDeviceStrategy(device='/gpu:0')
AttributeError: module 'tensorflow.contrib' has no attribute 'distribute'
Ok thanks, then it's the MirroredStrategy that I added recently that is only available from 1.8 on. I'll add a try-except for this.
Hi ! In case it could be interesting for reuse of your code, I made a Dockerfile (and added the documentation for running it in this context). I believe it could be much easier for people that already have some other version of cudas and the likes...