Unable to import soundfile from torchaudio

Hi there,
I am having issue when importing torchaudio.backend.soundfile_backend.load
Here is the full explanations:

  1. I clone my current working env to a new env in anaconda. I though everything should be working as usual as in my existing env. However, it is not.

  2. I run two of these:
    a) pip install soundfile
    b) conda install -c conda-forge pysoundfile

It returns:
a) Requirement already satisfied: soundfile in c:\users\user\anaconda3\envs\pytorch-gpu-teobased\lib\site-packages (0.11.0)
Requirement already satisfied: cffi>=1.0 in c:\users\user\anaconda3\envs\pytorch-gpu-teobased\lib\site-packages (from soundfile) (1.15.1)
Requirement already satisfied: pycparser in c:\users\user\anaconda3\envs\pytorch-gpu-teobased\lib\site-packages (from cffi>=1.0->soundfile) (2.21)

b) Collecting package metadata (current_repodata.json): done
Solving environment: done

==> WARNING: A newer version of conda exists. <==
current version: 22.9.0
latest version: 23.1.0

Please update conda by running

$ conda update -n base -c defaults conda

All requested packages already installed.

Retrieving notices: …working… done

  1. However, the error on jupyter notebook is still showing:
    RuntimeError: torchaudio.backend.soundfile_backend.load requires soundfile, but we failed to import it. Please check the installation of soundfile.

What should I do now? I also have rerun torchaudio conda reinstallation. But no luck.
Hope anyone can help me on this.
For cloning, do I actually need to reinstall pytorch, and others?


This SO thread does solve my issue: