Skip to content

Conversation

@patrickvonplaten
Copy link
Contributor

@patrickvonplaten patrickvonplaten commented Nov 13, 2025

What does this PR do?

This PR makes sure that:

from transformers import AutoTokenizer

tok = AutoTokenizer.from_pretrained("mistralai/Mistral-Small-3.2-24B-Instruct-2506")
tok = AutoTokenizer.from_pretrained("mistralai/Ministral-8B-Instruct-2410")

works correctly

@github-actions
Copy link
Contributor

[For maintainers] Suggested jobs to run (before merge)

run-slow: auto

@patrickvonplaten patrickvonplaten changed the title WIP Add AutoTokenizer mapping for mistral3 and ministral Nov 13, 2025
@patrickvonplaten
Copy link
Contributor Author

None of the test failures look related

@ArthurZucker ArthurZucker added the for patch Tag issues / labels that should be included in the next patch label Nov 14, 2025
@ArthurZucker ArthurZucker merged commit 7607d80 into huggingface:main Nov 14, 2025
18 of 21 checks passed
Copy link
Contributor

@vasqu vasqu left a comment

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

The auto mapping is incorrect and only works with mistral_common, see #41553 (comment). This is probably an issue for multiple mistral-related models now.

Tl;dr: It defines a llama backup but does not have the necessary files uploaded to the hub for that to work

@ArthurZucker
Copy link
Collaborator

Yep, but we need to sort that on our side iMO!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

for patch Tag issues / labels that should be included in the next patch

Projects

None yet

Development

Successfully merging this pull request may close these issues.

3 participants