Based on article https://arxiv.org/pdf/2001.04776.pdf
analysis.py- script used to create plots (distribution per generation) from log fileevolution.py- main script runs the evolutionary training algorithm distributed between several GPUsmodel_generation.py- module which generate model from list of units described onunitmoduletrain.py- training script which forming learning per each GPU, runs bysubprocessmodule fromevolutionscriptunit.py- module which working with binary structure of model (generation, transforming to units, transforming to hex/string)
foo@bar:~/DIP_NAS$ python evolution.py --img ./img/hr.png --gpus 4 |
 |
|---|
 |
 |
|---|
- PSNR distribution (mixin with random models)
- Implement losses for other task (in-paint, de-noise)