6
audio_codes = self.model.tts.dvae(gen_audio_mel_specs, mode='encode') 在执行上述代码时遇到了RuntimeError: Expected 2D (unbatched) or 3D (batched) input to conv1d, but got input of size: [1, 5, 512, 469]的问题 其中gen_audio_mel_specs的shape为torch.Size([5, 100, 938]) 代码片段为