Deep neural networks for system identification problems.
NOTE: This code implement the examples of the paper: Deep Convolutional Networks in System Identification. Better documentation and more readable code structure is a working in progress...
The code is possible to run in interactive mode mainly to enable easy access to evaluation
To get access to the model in interactive mode write
import run
(model, loaders, options) = run.run(load_model="logs/train_213214124/model.pt")
all_output = []
for i, (u, y) in enumerate(loaders["train"]):
all_output += [model(u,y)]
all_output = np.concat(all_output,0)