Dataset Viewer
The dataset viewer is not available because its heuristics could not detect any supported data files. You can try uploading some data files, or configuring the data files location manually.
🛠️ Vault Project: Offline Knowledge Vault
In a world of uncertainty, knowledge must endure.
Vault Project is a tool for creating an offline-optimized database of Wikipedia and other open knowledge repositories, ensuring accessibility even without an internet connection.
🌐 What is it?
Vault Project takes publicly available datasets (e.g., Wikipedia, OpenStreetMap, WikiHow) and structures them into an SQLite database with:
- Full-Text Search (FTS) for efficient offline queries.
- Embeddings for semantic search.
- Language-Specific Dumps for multiple Wikipedia editions.
🧠 Why Vault?
- Resilience: Access knowledge offline, even if online resources become unavailable.
- Efficiency: Indexed and optimized with embeddings and FTS for fast lookups.
- Customizable: Plug in custom datasets or personal documents.
📦 Dataset Structure
The dataset is organized by language:
/en/wikipedia_indexed.db- English Wikipedia/it/wikipedia_indexed.db- Italian Wikipedia/zh/wikipedia_indexed.db- Chinese Wikipedia...
Each database file includes:
- Articles with embeddings for semantic search.
- Full-text search (FTS5) on titles and content.
- Metadata tables for cross-referencing sources.
🛠️ Usage
Install dependencies:
pip install sqlite3 transformers huggingface_hub
Query the Database
import sqlite3
conn = sqlite3.connect("wikipedia_indexed.db")
cursor = conn.cursor()
query = "SELECT title, snippet FROM articles WHERE articles MATCH 'quantum physics';"
for title, snippet in cursor.execute(query):
print(f"{title}: {snippet}")
conn.close()
⚙️ Technical Insights
- SQLite for portability.
- FTS5 for full-text indexing.
- Embeddings generated with open-source NLP models.
📜 License
Distributed under the MIT License.
license: mit
- Downloads last month
- 33