IF icon indicating copy to clipboard operation
IF copied to clipboard

Published 20 hours ago • verified publisher icon deep-floyd

Flan-T5

Open Bachstelze opened this issue 1 year ago • 2 comments

Can we use FLAN-T5 as a language model? Those FLAN models can represent English and other languages significantly better in our tests. "If you already know T5, FLAN-T5 is just better at everything. For the same number of parameters, these models have been fine-tuned on more than 1000 additional tasks covering also more languages."

Bachstelze avatar May 03 '23 06:05 Bachstelze

Wouldn't you have to retrain text to image models because the representations are different?

phalexo avatar May 03 '23 12:05 phalexo

Yes, could expertiment with finetuning once you swap it out. I'm 1000% sure FLAN-T5 would result in higher fidelity output, better composition, way better spatial awareness. I think "tango model" kind of validates this.

darkman111a avatar May 03 '23 15:05 darkman111a