Hi everyone - for an text model, is there a simple way to de-tokenize the text from plot_top_losses()?
plot_top_losses()
It’s hard to read the text, sorry if it’s an obvious question, I’m struggling to figure it out.
