Model: Using 4 speaker Tino3141/4SpeakersDropCocktail180000 as pretraining, then trained on full dataset for 80000

5402f0b

verified

Tino3141 commited on Jan 30