What dataset is this model tuned on?

#1
by ljxw88 - opened

Hi! just want to know what dataset is this model tuned on? Thanks!

Owner

It's trained on a synthetic dataset where larger models such as Zephyr, Llama2 are prompted to generate positive continuation to toxic prefixes.
Will be publishing the dataset in the future.

jays changed discussion status to closed

Sign up or log in to comment