Contrastive learning seems interesting
A Dockerized agent that analyzes video clips and generates captions in four modes: formal, sarcastic, tech humor, and everyday humor.
.png&w=256&q=75)
