Transfer Learning 101: How AI Learns Faster and Smarter from Pre-Trained Models

Most people talk about “transfer” without really understanding what it means. Not everything you do in training is supposed to have a direct transfer to top end sprinting or sport performance and that’s fine. A lot of what we do is develop general outputs/capacities with hopes of it leading to improvements in the skill. We can classify transfer as positive, negative, or neutral, and direct or indirect. A-skips, dribbles, high knees, might not make you faster on their own, but they serve a purpos