Deeplearning framework for PyTorch
FOS is a Python framework that makes it easy to develop neural network models in PyTorch. Some of its main features are:
- Less boilerplate code required, see also the example below.
- Lightweight and no magic under the hood that might get in the way.
- You can extend Fos using common OO patterns.
- Get the insights you need into the performance of the model.
You can install FOS using pip:
pip install fos
Or alternatively from the source:
python setup.py install
Fos requires Python 3.5 or higher.
Training a model, requires just a few lines of code. First create the model, optimizer and loss function that you want to use, creating plain PyTorch objects:
net = resnet18() optim = Adam(predictor.parameters()) loss = F.binary_cross_entropy_with_logits
Then create the FOS objects that will take care of the training and output:
workout = Workout(predictor, loss, optim)
And we are ready to start the training:
trainer.fit(train_data, valid_data, epochs=5)
You can find several example Jupyter notebooks here, or even more convenient try them directly in a Google Colab environment:
- Basic Example
- MNIST example
If you want to help out, we appreciate all contributions. Please see the [contribution guidelines]() for more information.
As always, PRs are welcome :)=
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
|Filename, size||File type||Python version||Upload date||Hashes|
|Filename, size fos-0.7.0-py3-none-any.whl (21.0 kB)||File type Wheel||Python version py3||Upload date||Hashes View|
|Filename, size fos-0.7.0.tar.gz (16.1 kB)||File type Source||Python version None||Upload date||Hashes View|