huggingface_hub: Token cannot be pasted in JupyterNotebook in VScode

I was having the same issue in Jupyter and I was able to log in with: from huggingface_hub import interpreter_login

interpreter_login()

For anyone landing on this issue, here is a summary of how to login with the Hub:

The preferred way: use login

It will open the widget in a notebook and default to prompting the user in the terminal.

from huggingface_hub import login

login()

Without user input: login(token="hf_***")

It’s also possible to login by providing directly the token. For example, you can get it from an environment variable.

import os
from huggingface_hub import login

login(token=os.environ.get("HF_TOKEN"))

In this example, if the environment is set, it is used. If not, the user is prompted.

Force interpreter or notebook usage

Use interpreter_login or notebook_login if you want to force it to use 1 or the other (instead of the auto-detection). When doing this, you cannot provide the token as input.

For more details, here is the related package reference. Hope this can help future users 😃

As a temporary workaround, since you’re running on your personal machine and since the token is shared on a user-level, you should be able to do huggingface-cli login in your console and be logged-in in your notebook.

I was having the same issue in Jupyter and I was able to log in with: from huggingface_hub import interpreter_login

interpreter_login()

Thank you. This method worked for me.

I was having the same issue in Jupyter and I was able to log in with: from huggingface_hub import interpreter_login

interpreter_login()

thank you very much!

I was having the same issue in Jupyter and I was able to log in with: from huggingface_hub import interpreter_login

interpreter_login()

Thanks. I use vs code to login huggingface and apply ur approach. It worked for me.