ea56eb61f0
- Don't define training_net in define_optimizers - this drops the shell and leads to problems downstream - Get rid of support for multiple training nets per opt. This was half baked and needs a better solution if needed downstream. |
||
|---|---|---|
| .. | ||
| __init__.py | ||
| injectors.py | ||
| losses.py | ||
| progressive_zoom.py | ||
| steps.py | ||
| stereoscopic.py | ||
| tecogan_losses.py | ||