Opened this issue 2 years ago · 0 comments
It is mix of outputs in the code rather than mix of models in the paper. Why is it?