How do I use named backreference with re.sub?

I need to replace only full words so need to use backreference.
Re.sub works when using backreference:

import re

text = """
a=chef;
b=achefile;
"""

print(re.sub(r'([^\w])chef([^\w])', r'\1boss\2', text))

results (expectedly) in:

a=boss;
b=achefile;

However, what to do if I want to replace with a ‘0’?
The same regex breaks because the \1 is interpreted as \10…

print(re.sub(r'([^\w])chef([^\w])', r'\10\2', text))

I could not find how to use named groups with re.sub.
The search string would be:

r'(?P<start>[^\w])chef([^\w])'

but what will the replace string be?

>Solution :

Use \g<id>:

print(re.sub(r'([^\w])chef([^\w])', r'\g<1>0\g<2>', text))

# Output
a=0;
b=achefile;