#Model quality improvement
1 messages · Page 1 of 1 (latest)
I asked DR87, and he said no, and that it doesn't make sense architecturally-wise.
don't trust random redditor comments that don't make sense nor have proofs/explanations
1 messages · Page 1 of 1 (latest)
so on reddit i saw someone say that training your own voice and merging 10%~ with the model your using make the quality of the model better idk if it actually works or not can someone conferm or deny?
I asked DR87, and he said no, and that it doesn't make sense architecturally-wise.
don't trust random redditor comments that don't make sense nor have proofs/explanations