PyTorch is a python based library built to provide flexibility as a deep learning development platform. The workflow of PyTorch is as close as you can get to python’s scientific computing library – numpy.
Much like other frameworks, things like image recognition and timeseries sequences are a frequent use for PyTorch.
This particular model is using timeseries data to output a graph that can be viewed in your data bucket.
Applicable Scenarios and Problems
This model becomes an excellent demo on a few levels. The first one is that you can showcase how modules would be uploaded and added to create a PyTorch model here on the platform. You can talk through how a data scientist could create these modules outside and then push them to the platform as a way to deploy into production.
You can also showcase an option for data output on Algorithmia and showcase how the graph would dump out. Plus this is literally a demo of data for sales forecasting so it's a semi-business use case even if it is relatively simple.
Notes about this Demo
The graph output can be found in your data tab, shown in the photo below: