Nonnmy

Using 'python setup.py install' instead of pip3 install works for me.

> Hi, > > I had the same problem, but managed to fix it by forcing data type as float. The problem occurred when the input was Byte. > >...

In this issue #49 they meet very similar question in training time using their customized dataset. However, I just work with the provided code and data, and didn't change anything,

> I use https://github.com/peteanderson80/bottom-up-attention/ for feature extraction on my own images, and then run the image caption model, but the result caption is incomplete. > > e.g. >  >...