News for hackers. Stuff that matters.
Re:SuperPrompt: Better Text to Image Prompts in 77M Parameters by roborovskis on Thursday March 14, 2024 @04:07PM (#39708388)
Thanks for the kind words! I started with the 780M param flan-t5-large model, and kept trying smaller and smaller base models - I was shocked at how good the output was at 77M. As you go smaller, though, it's much easier to accidentally overfit or collapse the model and produce gibberish. Had to be very careful with hyperparams and sanitizing / filtering the dataset.

There are no comments on this comment yet.