Partha Ghosh

Perhaps something changed between python versions because of which memory is fragmented. Can you try using `.reshape()` as suggested?

Well, from a geometry perspective nothing really changes w.r.t resolution. So you can imagine all parameters except for the internal camera matrix everything stays the same. The only things that...

Please see the `First thing first` section of the read me. you will need several resource files, including the ones you mentioned.

you should only change the pose parameters of the flame code and hold the style vector and the rest of the parameters such as lighting and texture constant.

Sorry for the delay in the reply. Where exactly did this happen? Perhaps you are using latest tensorflow. This code was tested with tf 1.0

Still down. Also MNIST and others. Perhaps just make a dropbox or google drive link

@matthiasplappert can we put this through?

@matthiasplappert I think some part of this can be moved to keras itself. But some part of it has to be part of keras-rl as well. For example ddpg has...

How do you plan on getting the right conditioning?

Any update on this?