Autokeras: RuntimeError: invalid argument 2: input image (T: 4 H: 4 W: 1) smaller than kernel size (kT: 2 kH: 2 kW: 2)

[mmromero]

Hi,

I am training with data size [batches, 80, 80, 20, channels]. As of now I only have one channel but I will increase the number soon. My images are 3D volumes of 80x80x20 and I am using an ImageClassifier3D for a binary classification. Does anyone know what might be going on?

Thanks,
Miguel.

|               Training model 1               |
+----------------------------------------------+
Using TensorFlow backend.

No loss decrease after 5 epochs.


Saving model.
+--------------------------------------------------------------------------+
|        Model ID        |          Loss          |      Metric Value      |
+--------------------------------------------------------------------------+
|           1            |      0.6787109375      |   0.6288659793814433   |
+--------------------------------------------------------------------------+


+----------------------------------------------+
|               Training model 2               |
+----------------------------------------------+
Using TensorFlow backend.
Epoch-1, Current Metric - 0:   0%|                                        | 0/9 [00:00<?, ? batch/s]Process SpawnProcess-3:
Traceback (most recent call last):
  File "/home/miguel/anaconda3/envs/autokeras/lib/python3.6/multiprocessing/process.py", line 258, in _bootstrap
    self.run()
  File "/home/miguel/anaconda3/envs/autokeras/lib/python3.6/multiprocessing/process.py", line 93, in run
    self._target(*self._args, **self._kwargs)
  File "/home/miguel/anaconda3/envs/autokeras/lib/python3.6/site-packages/autokeras/search.py", line 351, in train
    raise e
  File "/home/miguel/anaconda3/envs/autokeras/lib/python3.6/site-packages/autokeras/search.py", line 344, in train
    verbose=verbose).train_model(**trainer_args)
  File "/home/miguel/anaconda3/envs/autokeras/lib/python3.6/site-packages/autokeras/nn/model_trainer.py", line 137, in train_model
    self._train()
  File "/home/miguel/anaconda3/envs/autokeras/lib/python3.6/site-packages/autokeras/nn/model_trainer.py", line 173, in _train
    outputs = self.model(inputs)
  File "/home/miguel/anaconda3/envs/autokeras/lib/python3.6/site-packages/torch/nn/modules/module.py", line 489, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/miguel/anaconda3/envs/autokeras/lib/python3.6/site-packages/autokeras/nn/graph.py", line 689, in forward
    temp_tensor = torch_layer(edge_input_tensor)
  File "/home/miguel/anaconda3/envs/autokeras/lib/python3.6/site-packages/torch/nn/modules/module.py", line 489, in __call__
    result = self.forward(*input, **kwargs)
  File "/home/miguel/anaconda3/envs/autokeras/lib/python3.6/site-packages/torch/nn/modules/pooling.py", line 638, in forward
    self.padding, self.ceil_mode, self.count_include_pad)
RuntimeError: invalid argument 2: input image (T: 4 H: 4 W: 1) smaller than kernel size (kT: 2 kH: 2 kW: 2) at /opt/conda/conda-bld/pytorch_1550802451070/work/aten/src/THCUNN/generic/VolumetricAveragePooling.cu:57

[darren1231]

I have the same problem. Can somebody explain why??

[stale[bot]]

This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.