Question
#1
by
isr431
- opened
From your personal experience, how does this new version with both datasets compare to the previous one? Sorry to trouble you, but could you also make an updated finetune of Lyra Gutenberg (based on Lyra v1, not v4) as it performs better than the Lyra v4 finetune.
I honestly haven't done extensive testing with the new dataset. But yes, I can queue a Lyra v1 tune as well.