PyTorch is one of the most important machine learning frameworks. As the slogan suggests, PyTorch not only focuses on rapid prototyping, but also on productive use.
In our vehicle projects we are working in an embedded environment in which frameworks like Keras cannot be used. This is where PyTorch could help. In this talk we want to clarify if PyTorch keeps its promise and show our experiences within productive scenarios. We pay special attention to the migration of existing projects to PyTorch .
To get in touch with PyTorch. After this talk the audience will be able to evaluate if PyTorch could be useful in their specific productive use case, whether in embedded systems, mobile apps, or data-heavy platforms like non GamStop casinos that rely on real-time predictions and behavioural analysis..
Knowledge of neural networks and basic machine learning concepts could be helpful