README.md

metadata

title: Embedding Atlas Explorer
emoji: 🗺️
colorFrom: blue
colorTo: green
sdk: gradio
sdk_version: 5.42.0
app_file: app.py
pinned: true

Embedding Atlas Explorer

This Space allows you to visualize any text-based dataset from the Hugging Face Hub using Apple's Embedding Atlas.

How it works:

1. Load Data: Enter the name of a dataset from the Hub, specify the text column, and choose a sample size.
2. Generate Embeddings: The backend uses a Sentence Transformer model to convert the text into numerical vectors (embeddings).
3. Reduce Dimensions: UMAP is used to project the high-dimensional embeddings down to a 2D map for visualization.
4. Explore: The interactive Atlas UI is displayed, allowing you to pan, zoom, select points, and explore the structure of your text data.