search

modelscope / KAN-TTS

KAN-TTS is a speech-synthesis training framework, please try the demos we have posted at https://modelscope.cn/models?page=1&tasks=text-to-speech
MIT License
493 stars 79 forks source link

train_sambert报错 #69

Open yw1991 opened 1 year ago

yw1991 commented 1 year ago

inputs_text_embedding + pitch_embeddings + energy_embeddings RuntimeError: The size of tensor a (50) must match the size of tensor b (411) at non-singleton dimension 1 2023-07-03:11:07:19 INFO [trainer.py:903] torch.Size([32, 50, 4]) 2023-07-03:11:07:19 INFO [trainer.py:904] torch.Size([32, 50]) 2023-07-03:11:07:19 INFO [trainer.py:905] torch.Size([32, 50]) torch.Size([32, 50, 32]) torch.Size([32, 411, 32]) torch.Size([32, 411, 32])