Aquila DB is a Neural search engine. In other words, it is a database to index Latent Vectors generated by ML models along with JSON Metadata to perform k-NN retrieval. It is dead simple to set up, language-agnostic, and drop in addition to your Machine Learning Applications. Aquila DB, as of current features is a ready solution for Machine Learning engineers and Data scientists to build Neural Information Retrieval applications out of the box with minimal dependencies.
This project is still in alpha version & we're already using it in production to power semantic search at https://aquila.network.
Wanna support this project? Yes, we love getting a star ⭐ and shout-out 🗣️ 🤗
Join Community chat and get support:
- If you are working on a data science project and need to store a hell of a lot of data and retrieve similar data based on some feature vector, this will be a useful tool to you, with extra benefits a real world web application needs.
- Are you dealing with a lot of images and related metadata? Want to find similar ones? You are at the right place.
- If you are looking for a document database, this is not the right place for you.
Aquila DB powers search features of Aquila Network. Here is where Aquila DB fits in the entire ecosystem:
If you are serious and wanna dive down the rabbit hole, read our whitepapers and technical specifications (being actively worked on).
As a side note, everything in Aquila Network is defined by the specifications and a large chunk of our efforts goes into it. We also maintain quality implementations of those specifications with non-technical users in mind. This is to make sure that - Aquila Network is fully open, decentralized by design, and Fair. You can follow those specifications to implement your alternative software and still interact with the network without any restrictions.
Run curl -s -L https://raw.githubusercontent.com/Aquila-Network/AquilaDB/master/install.sh | /bin/bash -s -- -d 1 .
You need docker installed in your system
Build image (lite): docker build https://raw.githubusercontent.com/Aquila-Network/AquilaDB/master/Dockerfile -t aquiladb:local
Build image (big data): docker build https://raw.githubusercontent.com/Aquila-Network/AquilaDB/master/DockerfileBig -t aquiladb:localbg
Run image (to deploy Aquila DB lite): docker run -p 5001:5001 -d aquiladb:local
Run image (to deploy Aquila DB big): docker run -p 5001:5001 -d aquiladb:localbg
We currently have multiple client libraries in progress to abstract the communication between deployed Aquila DB and your applications.
When you use a client library to authenticate with AquilaDB, you might need access the same private key (wallet key) used by AquilaDB. This key is located inside /ossl/ directory within AquilaDB docker container (in your computer if you have installed AquilaDB directly without docker). To access the keys inside your AquilaDB container, follow below steps:
- identify
CONTAINER IDfor the already runningaquiladbdocker instance:docker ps - take a copy of private keys from docker container to your host machine:
docker cp CONTAINER_ID:/ossl/ ./ - now you will see a new directory named
osslat your current location. Use the keys inside it.
If your pipeline requires the private keys to be generated in advance, you can do it in your host machine and then mount it to the container's /ossl/ directory.
Run:
mkdir -p <host>/ossl/
openssl genrsa -passout pass:1234 -des3 -out <host>/ossl/private.pem 2048
openssl rsa -passin pass:1234 -in <host>/ossl/private.pem -outform PEM -pubout -out <host>/ossl/public.pem
openssl rsa -passin pass:1234 -in <host>/ossl/private.pem -out <host>/ossl/private_unencrypted.pem -outform PEM
This project is still and will be under active development with intermediate production releases. It can either be used as a standalone database or as a participating node in Aquila Network. Please note, Aquila Port (peer-peer network layer for Aquila DB nodes) is also a work in progress. Currently, you need to deploy your custom models to feed vector embeddings to Aquila DB, until Aquila Hub developments get started.
We have prepared a document to get anyone interested to contribute, immediately started with Aquila DB. Here is our high-level release roadmap.
We have started meeting developers and do small talks on Aquila DB. Here are the slides that we use on those occasions: http://bit.ly/AquilaDB-slides
Video:
As of current AquilaDB release features, you can build Neural Information Retrieval applications out of the box without any external dependencies. Here are some useful links to learn more about it and start building:
- Microsoft published a paper and youtube video on this to onboard anyone interested:
- Embeddings for Everything: Search in the Neural Network Era: https://www.youtube.com/watch?v=JGHVJXP9NHw
- Autoencoders are one such deep learning algorithms that will help you to build semantic vectors - foundation for Neural Information retrieval. Here are some links to Autoencoders based IR:
- Note that, the idea of information retrieval applies not only to text data but for any data. All you need to do is, encode any source datatype to a dense vector with deep neural networks.
email us to sponsor this project [email protected].
If you use Aquila DB in an academic paper, we would 😍 to be cited. Here are the two ways of citing Aquila DB:
\footnote{https://github.com/Aquila-Network/AquilaDB}
@misc{AquilaNetwork2019AquilaDB,
title={AquilaDB: Neural Search Engine},
author={Jubin Jose, Nibin Peter},
howpublished={\url{https://github.com/Aquila-Network/AquilaDB}},
year={2019}
}
Apache License 2.0 license file
created by ❤️ with a-mma (a_മ്മ)