A playbook for systematically maximizing the performance of deep learning models.
High-quality implementations of standard and SOTA methods on a variety of tasks.
Open on GitHub