RuntimeError: stack expects a non-empty TensorList

[noahschulman14]

Hello, I am getting the following error when trying to use tortoise. Everything used to work fine for me.

Here is the stack trace:

Traceback (most recent call last): File "/mnt/MFC_deepfake_data/ntest/tortoise/tortoise-stuff/tortoise-tts/tts.py", line 36, in gen = tts.tts_with_preset(text, voice_samples=voice_samples, conditioning_latents=conditioning_latents, preset=preset) File "/mnt/MFC_deepfake_data/ntest/tortoise/tortoise-stuff/tortoise-tts/tortoise/api.py", line 328, in tts_with_preset return self.tts(text, **settings) File "/mnt/MFC_deepfake_data/ntest/tortoise/tortoise-stuff/tortoise-tts/tortoise/api.py", line 393, in tts auto_conditioning, diffusion_conditioning, autoconds, = self.get_conditioning_latents(voice_samples, return_mels=True) File "/mnt/MFC_deepfake_data/ntest/tortoise/tortoise-stuff/tortoise-tts/tortoise/api.py", line 274, in get_conditioning_latents auto_conds = torch.stack(auto_conds, dim=1) RuntimeError: stack expects a non-empty TensorList

Any help would be much appreciated!!!

[riazkhawaja]

Are you using a .wav file?

[Zeon8985Army]

In my condition, my reference voice is empty