In this issue: we start the new series about text-to-image models; we discuss CLIP, a neural network that can learn image representations while being trained using natural language datasets; we explore Hugging Face’s CLIP implementation. Enjoy the learning!