I'm going to have to wait to generate this one, the model I used for Gabe Newell is super slow, it would take hours to output the whole 2 minutes worth with different voices. Currently trying to train Microsoft's VALLE-E which hopefully will be both faster and produce better results.
I'm going to have to wait to generate this one, the model I used for Gabe Newell is super slow, it would take hours to output the whole 2 minutes worth with different voices. Currently trying to train Microsoft's VALLE-E which hopefully will be both faster and produce better results.
(post is archived)