so-vits-svc/vencoder/encoder.py

12 lines
358 B
Python

class SpeechEncoder(object):
def __init__(self,vec_path = "pretrain/checkpoint_best_legacy_500.pt"):
self.model = None #This is Model
self.hidden_dim = 768
pass
def encoder(self,wav):
'''
input: wav:[batchsize,signal_length]
output: embedding:[batchsize,wav_frame,hidden_dim]
'''
pass