[Docs] Standardize ConvNeXt V2 model card

[CYBki]

What does this PR do?

This PR standardizes the ConvNeXt V2 model card to match the Hugging Face model card template described in #36979.

Changes

• Added standardized header and badges.
• Wrote a short, user-friendly model description explaining what makes ConvNeXt V2 unique.
• Added Pipeline, AutoModel, and transformers-cli usage examples.
• Included quantization snippet using BitsAndBytesConfig.
• Documented intended uses, limitations, preprocessing details, and model variants.
• Linked resources (paper, implementation, example scripts, Colab notebook).
• Added citation section.

Notes

• ConvNeXt V2 is primarily Conv2d, so quantization benefits are modest (mainly the classifier head).
• Evaluation accuracy is sensitive to input resolution (224 vs 384).

Checklist

• Matches model card standardization template.
• Doc-builder syntax preserved (hfoptions, [!TIP], [[autodoc]]).
• Examples tested with facebook/convnextv2-tiny-1k-224.

---

Who can review?

Documentation: @stevhliu

[CYBki]

Hi! This is a documentation-only change (model card for ConvNeXt V2).
All required checks are green ✅

Could a maintainer please approve the docs workflow and review?
cc @stevhliu

[stevhliu]

Thank you so much for the time and effort you put into this—I really appreciate it. With the number of models in Transformers growing so quickly, I’ve been looking into ways to automate model card generation to keep up with the pace. For now, I’ll be closing this PR, but I’m very grateful for your interest in this project!

Thanks again 🤗