Fix model card: base model, title, and image-text-to-text quick start

#3
by SamJoshua - opened

Hi! The auto-generated card shows "Model Card for None", a base model of None, and a text-only pipeline("text-generation", model="None") quick start, which doesn't work for this image-text-to-text model.

This PR:

  • Sets base_model: HuggingFaceTB/SmolVLM2-2.2B-Instruct and pipeline_tag: image-text-to-text, and adds gui-agent / computer-use / smolagents tags
  • Replaces the title and intro with a short description of the two-phase Smol2Operator recipe, based on the Smol2Operator blog post
  • Replaces the quick start with an AutoModelForImageTextToText screenshot + instruction example, pointing to the blog and recipe notebook for the exact action format
  • Removes the licence: license placeholder typo (I left the license unset, since there's an open discussion asking about it)

Training details, W&B link, framework versions and citation are unchanged. Happy to adjust anything!

Ready to merge
This branch is ready to get merged automatically.

Sign up or log in to comment