README.md

Image Classification — Gradio + Hugging Face

This Space wraps a pretrained image classifier (Vision Transformer) with a simple Gradio UI.

How to use

1. Upload a photo (PNG/JPG).
2. Choose how many classes to display (Top‑K).
3. Click Classify to see the predicted labels with confidence scores.

Model

• Default: google/vit-base-patch16-224 (ImageNet‑1k)
• Alternatives that work on CPU Basic:
  • microsoft/resnet-50
  • facebook/convnext-tiny-224

Notes: The classifier is trained on ImageNet‑1k; some objects (logos, memes, niche items) may not map cleanly to labels.