A visualisations library for PyTorch with torchbearer
Torchbearer.visual is a companion package to torchbearer which is intended to re-implement state of the art practices in filter and feature visualisation.
|
Quickstart: Our quickstart guide, go here to learn the basics of visualising deep networks with visual. |
|
|
||
|
||
|
Compositional Pattern Producing Networks: A quick guide on using the CPPNImage class in visual to create a very different kind of visualisation. |
|
|
|
||
|
|
||
|
Deep Dream: A guide on performing deep dreams with visual, including pyramid ascent in Fourier space. |
|
|
|
||
|
|
||
| Custom Models: How to use visual with your own model or one from torch.hub. |
|
|
|
|
||
|
|
||
The easiest way to install visual is with pip:
pip install torchbearer-visual
Alternatively, build from source with:
pip install git+https://github.com/pytorchbearer/visual