PyTorch Introduction — Using Custom Data
In this post of the PyTorch Introduction, we’ll learn how to use custom datasets with PyTorch, particularly tabular, vision and text data
PyTorch is one of the hottest libraries in the Deep Learning field right now. Since ChatGPT’s release, deep learning libraries have arguably garnered the most attention among data scientists and machine learning engineers, particularly due to the current practical applications they enable.
With their extensive capability of performing complex multidimensional calculations extremely fast, these libraries changed the way we train Neural Network models. Particularly, they are extremely helpful in managing the large number of weights that these models store and optimize. Rivaling with TensorFlow (Google’s framework), PyTorch is Meta’s open-source framework, giving you the ability to train deep learning models with a very cool and practical syntax.
So far, in this PyTorch series, we’ve learned a couple of fundamentals that gave us the ability to work with this library. For example:
