Errata

Hi,

I have a question how to run the code in Chapter 3.
About the audio models ,examples 3-4 , 3-5 and 3-6.py.
In 3-4.py we have
# schemas.py

from typing import Literal

VoicePresets = Literal["v2/en_speaker_1", "v2/en_speaker_9"]
And
# models.py

import torch
import numpy as np
from transformers import AutoProcessor, AutoModel, BarkProcessor, BarkModel
from schemas import VoicePresets

device = torch.device("cuda" if torch.cuda.is_available() else "cpu")

I create from 3-4.py 2 files schemas.py and models.py . Is this the right way to proceed ?
In models.py , there should be from schemas import VoicePresets ;

I proceed the same way in 3-5.py and 3-6.py
To launch the FastAPI app, I do fastapi dev main.py (is that correct ?) but get an import error :"ImportError: cannot import name 'VoicePresets' from 'schemas'

Could you please provide some advice on how to run the code on the githup repo.
Where can I get access to a forum with FAQ for each chapter please?

Note from the Author or Editor:
You’re mostly on the right track, but here’s how to fix the ImportError and run the code properly.

You’re seeing:

ImportError: cannot import name 'VoicePresets' from 'schemas'

This is likely because:

Python is importing a different schemas.py (e.g. from another installed package).
Your project structure may not be correct (Python might not find the right schemas.py)

How to Fix

Make sure your directory looks like this:

project/
│
├── main.py
├── models.py
├── schemas.py
├── utils.py

All four files should be in the same folder.

2. Check your import paths

In models.py, this is correct:

from schemas import VoicePresets

As long as schemas.py is in the same folder, that should work.

If it’s in a subfolder like src/schemas.py, then you’d need:

from src.schemas import VoicePresets

Make sure there’s no other schemas.py shadowing it (check for naming conflicts in your virtual environment or system path).

3. You can also try using the following command to start the app

uvicorn main:app --reload

Make sure you’re in the same directory as main.py.

You need to have uvicorn installed:

pip install uvicorn


Optional: Use an __init__.py

To make your directory behave more like a Python module (especially if using subfolders), you can add an empty __init__.py file:

project/
├── __init__.py
├── main.py
├── models.py
├── schemas.py
├── utils.py


About FAQs or forum for the book

The best place to handle FAQs and community questions is to use GitHub Issues or Discussions tab on the repo

A simple regex pattern that captures the URLs into a named group called url and
matches both http and https protocols.

-----------------------------------
Should we revise“url”to urls”

Note from the Author or Editor:
Yes, please, should be "urls"

Instead of returning a chunk, yield it so that the load() function becomes an
asynchronous generator. Asynchronous generators can be iterated with async
for loops so that blocking operations within them can be awaited to let the
event loop start/resume other tasks.

--------------------
"async" and "loops: areis just text, not code, right?

Note from the Author or Editor:
"async for" should be code font, but "loops" is regular text.

For this example, you need to install the pytest-mocks plug-in,
which is a thin wrapper over Python’s built-in mocks library to
simplify mocking implementation. Use the following command to
install pytest-mocks:
--------------------------
Should we revise“pytest-mocks”to “pytest-mock”? Thank you.

Note from the Author or Editor:
Yes, please, should be "pytest-mock"