Apppublic

svetfm/epstein-fbi-rag

sourceHugging Facecc-by-4.0updated 9mo agoView on Hugging Face

0likes

App README

FBI Epstein Files - Semantic Search

Search through 236K document chunks from the FBI's release of Jeffrey Epstein-related files.

Features

—Semantic Search: Find documents by keyword or phrase
—Similar Documents: Find documents similar to a given Bates number
—Statistics: View collection metadata

Dataset

Embeddings loaded from svetfm/epstein-fbi-files

—236,174 chunks with 768-dimensional embeddings
—6,618 unique documents
—Model: nomic-embed-text via Ollama

API Usage

python

import gradio_client

client = gradio_client.Client("svetfm/epstein-fbi-rag")

# Search
results = client.predict(query="maxwell", top_k=10, api_name="/search")

# Find similar
similar = client.predict(doc_id="EFTA00000001", top_k=10, api_name="/find_similar")

# Get stats
stats = client.predict(api_name="/get_stats")

License

Public domain government documents. Dataset compilation CC-BY-4.0.