Guest
Data Fest Online 2020 Catalyst Workshop Track You probably have seen a lot of beginners in the field of deep learning struggle to write reusable well-structured code in their projects. Around a year ago I was at that point as well. Luckily I bumped on a Kaggle kernel which was using a Catalyst framework and it really encouraged me to dive deeper in Catalyst and apply its best practices in my own projects. Very soon I earned a bronze medal in semantic segmentation competition. After the competition has ended I decided to write a small research paper describing my approach. This allowed me to qualify into two machine learning summer schools this year. At the same time I was getting better in Catalyst and DL and won a silver medal in another computer vision task. After the finish of the second competition I was able to contribute to Catalyst library by implementing a callback which applies a Cutmix data augmentation. As a result, not only I learned how to write deep learning pipelines, but also built projects which allowed me to participate in different academic events and get noticeable lines in my CV. Register and get access to the tracks: Join the community:
