Update README

README.md

@@ -1,15 +1,47 @@
 ---
-title: Mnistvit
-emoji: 👁
-colorFrom: red
+title: mnistvit MNIST Vision Transformer
+emoji: 🔢
+colorFrom: yellow
 colorTo: gray
 sdk: gradio
-sdk_version: 6.25.0
+sdk_version: 6.24.0
 python_version: '3.12'
 app_file: app.py
 pinned: false
 license: mit
 short_description: Classify digits in MNIST-like images
+tags:
+- mnist
+- vit
+- vision
+- transformer
+- classification
+- recognition
+- image
+- digit
+models:
+- asnelt/mnistvit
+datasets:
+- ylecun/mnist
+thumbnail: >-
+  https://huggingface.co/spaces/asnelt/mnistvit/resolve/main/thumbnail.png
 ---
+# mnistvit - MNIST Digit Classification
 
-Check out the configuration reference at https://huggingface.co/docs/hub/spaces-config-reference
+This Space showcases mnistvit, a vision transformer (ViT) trained on MNIST with
+a PyTorch-only implementation, achieving 99.65% test set accuracy.
+
+## Model Details
+
+- **Hugging Face:**
+  [https://huggingface.co/asnelt/mnistvit/](https://huggingface.co/asnelt/mnistvit/)
+- **Code Repository:**
+  [https://github.com/asnelt/mnistvit/](https://github.com/asnelt/mnistvit/)
+- **Python Package Index:**
+  [https://pypi.org/project/mnistvit/](https://pypi.org/project/mnistvit/)
+
+## Use
+
+Draw a digit, or upload an image of one, then click the **Submit** button to see
+whether mnistvit recognizes the digit.  Support for MCP tools and agent use is
+provided as well.