rogertrullo/pytorch_convlstm

convolutional lstm implementation in pytorch

pytorch_convlstm

This is my attempt to implement convolutional lstm in pytorch. This idea has been proposed in this paper: Convolutional LSTM Network: A Machine Learning Approach for Precipitation Nowcasting

It's still in progress..

pytorch_convlstm's Issues

Is code ready to use?

If the code is ready to be used then please update readme file on how to use it.

TypeError: 'float' object cannot be interpreted as an integer

When I run the code, there's a TypeError in the model's forward part which in code is line 134
The error is

"TypeError: 'float' object cannot be interpreted as an integer".

Has anyone meet the same problem and fixed it?
I am now using pytorch 1.0.1 and python 3.6

RuntimeError

Given groups=1, weight of size 40 13 5 5, expected input[10, 20, 25, 25] to have 13 channels, but got 20 channels instead

Any news?

Hi , any news for this implementation? I think 'wight_init' has to be defined