error while training

i am trying to run your model, for that i have created a GRID folder in the data directory as mentioned in the repository with the same structure. but I am just taking 10 videos and their alignments.

When i run the train_multiuser file it is showing

Epoch 1/500

UnboundLocalError Traceback (most recent call last)
in ()
5 #validation_data=val_gen, nb_val_samples=nb_val_samples,
6 #callbacks=[checkpointer, csv, statisticCallback],
----> 7 callbacks=[statisticCallback, statisticCallback_unseen]
8 )

/usr/local/lib/python3.5/dist-packages/keras/legacy/interfaces.py in wrapper(*args, **kwargs)
86 warnings.warn('Update your ' + object_name + 87 ' call to the Keras 2 API: ' + signature, stacklevel=2)
---> 88 return func(*args, **kwargs)
89 wrapper._legacy_support_signature = inspect.getargspec(func)
90 return wrapper

/usr/local/lib/python3.5/dist-packages/keras/engine/training.py in fit_generator(self, generator, steps_per_epoch, epochs, verbose, callbacks, validation_data, validation_steps, class_weight, max_q_size, workers, pickle_safe, initial_epoch)
1911 epoch_logs['val_' + l] = o
1912
-> 1913 callbacks.on_epoch_end(epoch, epoch_logs)
1914 epoch += 1
1915 if callback_model.stop_training:

UnboundLocalError: local variable 'epoch_logs' referenced before assignment

PLEASE HELP!

[klauscc]

The problem may be the version incompatibles of keras and tensorflow. The tensorflow version is 1.4.0 and Keras version is 2.0.1. Moreover, runing the code under python3 env is not tested.