Python is very well known for its ecosystem of mature scientific computing packages. They range from low-level providers of generic functionality like NumPy to very domain specific tools like scikit-learn. A few years back, machine learning frameworks could be easily classified into the second category, but this is changing now. They become increasingly powerful and start to be applied in a whole variety of settings different than their original purpose. In this talk, I’ll present a basic tour of PyTorch, a relatively new library that has already gathered a significant user base. I’ll describe features useful both in machine learning and other domains, explain how it fits into the Python landscape, and showcase scenarios where it provides benefits over existing tools.
in __on domenica 22 aprile at 09:45 **See schedule**