forked from ecker/DL-Art-School
Allows downsampling immediately before processing, which reduces network complexity on higher resolution images but keeps a higher filter count. |
||
|---|---|---|
| .. | ||
| archs | ||
| __init__.py | ||
| base_model.py | ||
| loss.py | ||
| lr_scheduler.py | ||
| networks.py | ||
| SR_model.py | ||
| SRGAN_model.py | ||