|
|
@ -491,7 +491,7 @@ class TTSExecutor(BaseExecutor):
|
|
|
|
# multi speaker
|
|
|
|
# multi speaker
|
|
|
|
if am_dataset in {'aishell3', 'vctk', 'mix', 'canton'}:
|
|
|
|
if am_dataset in {'aishell3', 'vctk', 'mix', 'canton'}:
|
|
|
|
mel = self.am_inference(
|
|
|
|
mel = self.am_inference(
|
|
|
|
part_phone_ids, spk_id=paddle.to_tensor(spk_id))
|
|
|
|
part_phone_ids, spk_id=paddle.to_tensor([spk_id]))
|
|
|
|
else:
|
|
|
|
else:
|
|
|
|
mel = self.am_inference(part_phone_ids)
|
|
|
|
mel = self.am_inference(part_phone_ids)
|
|
|
|
self.am_time += (time.time() - am_st)
|
|
|
|
self.am_time += (time.time() - am_st)
|
|
|
|