ValueError: If printing histograms, validation_data must be provided, and cannot be a generator.

Hi, Thanks for your software.After creating data as test,there is an issue when i execute dcpg_train.py ./data/c1_000000-001000.h5 --val_files ./data/c13_000000-001000.h5 --cpg_model RnnL1 --out_dir ./models/cpg --nb_epoch 1 --nb_train_sample 1000 --nb_val_sample 1000 Here is the traceback: Traceback (most recent call last): File “/home/ljc/down/deepcpg-keras2/scripts/dcpg_train.py”, line 832, in <module> app.run(sys.argv) File “/home/ljc/down/deepcpg-keras2/scripts/dcpg_train.py”, line 194, in run return self.main(name, opts) File “/home/ljc/down/deepcpg-keras2/scripts/dcpg_train.py”, line 802, in main verbose=0) File “/home/ljc/anaconda3/lib/python3.6/site-packages/keras/legacy/interfaces.py”, line 87, in wrapper return func(*args, **kwargs) File “/home/ljc/anaconda3/lib/python3.6/site-packages/keras/engine/training.py”, line 2082, in fit_generator callbacks.on_epoch_end(epoch, epoch_logs) File “/home/ljc/anaconda3/lib/python3.6/site-packages/keras/callbacks.py”, line 77, in on_epoch_end callback.on_epoch_end(epoch, logs) File “/home/ljc/anaconda3/lib/python3.6/site-packages/keras/callbacks.py”, line 751, in on_epoch_end raise ValueError('If printing histograms, validation_data must be ’ ValueError: If printing histograms, validation_data must be provided, and cannot be a generator.

Could you plz help me?Thanks a lot.

[cangermueller]

I added a flag ‘–no_tensorboard’ to dcpg_traing.py to disable Tensorboard summaries during training. Can you update your code (git pull) and check if it works when you are using dcpg_train.py --no_tensorboard?

[zao]

I recommend using Tensorboard instead of Theano since Theano is dead.

Did you mean TensorFlow here? I must admit that I’m not entirely up to speed with the backends and tools around this field, as I’m a cluster admin, not a user.

My understanding and concern is that if you use ‘tensorflow’ as the Keras backend, DeepCpG just won’t work unless you supply the --no_tensorboard parameter. Is there any scenario where that actually works and shouldn’t just be automatically disabled? Is there any value to be gained from the TensorBoard functionality, or should the builtin output from DeepCpG be quite sufficient to do science with?

As it currently stands with DeepCpG 1.0.5, if the user or the site admin has configured the Keras backend to be TensorFlow, they’re in for an automatic batch queue failure unless they happen to know to pass in --no_tensorboard.