3D training error

[RenieWell]

I am try to run the notebook in \3D\examples\01_traing.pynb. The code runs good in Epoch 1, but it fails in Epoch 2, and shows that

tensorflow.python.framework.errors_impl.InvalidArgumentError: You must feed a value for placeholder tensor 'Placeholder' with dtype float and shape [?,?,?,?,1] [[{{node Placeholder}} = Placeholder[dtype=DT_FLOAT, shape=[?,?,?,?,1], _device="/job:localhost/replica:0/task:0/device:GPU:0"]()]] [[{{node lambda_2/clip_by_value/_809}} = _Recv[client_terminated=false, recv_device="/job:localhost/replica:0/task:0/device:CPU:0", send_device="/job:localhost/replica:0/task:0/device:GPU:0", send_device_incarnation=1, tensor_name="edge_451_lambda_2/clip_by_value", tensor_type=DT_FLOAT, _device="/job:localhost/replica:0/task:0/device:CPU:0"]()]]

It seems is a error in n2v_standard.py in line 216

history = self.keras_model.fit_generator(generator=training_data, validation_data=(validation_X, validation_Y),epochs=epochs, steps_per_epoch=steps_per_epoch, callbacks=self.callbacks, verbose=1)

Can you help me with this? BTW, my tensorflow version is 1.12.0 keras version is 2.2.1 numpy version is 1.16.3

[tibuch]

Hi @RenieWell

It might be that you are using an old N2V version. The current release is 0.1.4. Which version do you have installed?

[CMCI]

Hi @tibuch I am also getting the same error, having just used git pull master to get the most up-to-date version of the code. It also occurs after having run through epoch 1, but before reaching epoch 2.

[tibuch]

Hi @RenieWell

Unfortunately we had the wrong version numbers on our master-branch. I merged the correct version numbers and now everything should work again.

Sorry about the inconvenience and thank you for reporting the issue!

[RenieWell]

It works! Thanks a lot!