lucidrains / naturalspeech2-pytorch

Implementation of Natural Speech 2, Zero-shot Speech and Singing Synthesizer, in Pytorch
MIT License
1.26k stars 100 forks source link

The package fails to reference naturalspeech2_pytorch.utils.* #18

Open ArEnSc opened 1 year ago

ArEnSc commented 1 year ago

^ on an related @lucidrains is this ready for testing? I am thinking about trying to get it to memorize a sample to test and then pretrain a model. I did notice that you mentioned that the pitch conditioning wasn't complete

lucidrains commented 1 year ago

@ArEnSc hmm, can you show me the full stack trace?

ArEnSc commented 1 year ago

@lucidrains

@ArEnSc hmm, can you show me the full stack trace?

│ /usr/local/lib/python3.10/dist-packages/IPython/core/interactiveshell.py:3553 in runcode │ │ │ │ 3550 │ │ │ │ elif async : │ │ 3551 │ │ │ │ │ await eval(code_obj, self.user_global_ns, self.user_ns) │ │ 3552 │ │ │ │ else: │ │ ❱ 3553 │ │ │ │ │ exec(code_obj, self.user_global_ns, self.user_ns) │ │ 3554 │ │ │ finally: │ │ 3555 │ │ │ │ # Reset our crash handler in place │ │ 3556 │ │ │ │ sys.excepthook = old_excepthook │ │ in <cell line: 2>:2 │ │ │ │ /usr/local/lib/python3.10/dist-packages/naturalspeech2_pytorch/init.py:8 in │ │ │ │ 5 │ from einops._torch_specific import allow_ops_in_compiled_graph │ │ 6 │ allow_ops_in_compiled_graph() │ │ 7 │ │ ❱ 8 from naturalspeech2_pytorch.naturalspeech2_pytorch import ( │ │ 9 │ NaturalSpeech2, │ │ 10 │ Transformer, │ │ 11 │ Wavenet, │ │ │ │ /usr/local/lib/python3.10/dist-packages/naturalspeech2_pytorch/naturalspeech2_pytorch.py:27 in │ │ │ │ │ │ 24 from beartype.door import is_bearable │ │ 25 │ │ 26 from naturalspeech2_pytorch.attend import Attend │ │ ❱ 27 from naturalspeech2_pytorch.utils.tokenizer import Tokenizer, ESpeak │ │ 28 from naturalspeech2_pytorch.version import version │ │ 29 │ │ 30 from accelerate import Accelerator │ ╰──────────────────────────────────────────────────────────────────────────────────────────────────╯ ModuleNotFoundError: No module named 'naturalspeech2_pytorch.utils.tokenizer'

reproduce !pip install naturalspeech2-pytorch
then paste in

import torch from naturalspeech2_pytorch import ( EncodecWrapper, Model, NaturalSpeech2 )

lucidrains commented 1 year ago

@ArEnSc would you like to retry with 0.0.38?

ArEnSc commented 1 year ago

@lucidrains it works! Quick question so is it ready for training minus the pitch conditioning ?