Use late-interaction multi-modal models such as ColPALI in just a few lines of code.
Project description
Welcome to Byaldi
Did you know? In the movie RAGatouille, the dish Remy makes is not actually a ratatouille, but a refined version of the dish called "Confit Byaldi".
Byaldi is RAGatouille's mini sister project. It is a simple wrapper around the ColPali repository to make it easy to use late-interaction multi-modal models such as ColPALI with a familiar API.
Getting started
First, a warning: This is a pre-release library, using uncompressed indexes and lacking other kinds of refinements. Eventually, we'll add an HNSW indexing mechanism, pooling, and, who knows, maybe 2-bit quantization?
IIt'll get updated as the multi-modal ecosystem develops further!
Pre-requisites
ColPali access
ColPali is currently the only model of its kind. As it is based on PaliGemma, you will need to accept Google's license agreement for PaliGemma on HuggingFace, and use your own HF token to download the model.
Poppler
To convert pdf to images with a friendly license, we use the pdf2image
library. This library requires poppler
to be installed on your system. Poppler is very easy to install by following the instructions on their website. The tl;dr is:
MacOS with homebrew
brew install poppler
Debian/Ubuntu
sudo apt-get install -y poppler-utils
Project details
Release history Release notifications | RSS feed
Download files
Download the file for your platform. If you're not sure which to choose, learn more about installing packages.
Source Distribution
Built Distribution
Hashes for Byaldi-0.0.1rc1-py3-none-any.whl
Algorithm | Hash digest | |
---|---|---|
SHA256 | 3e7a0f42ef094d3b9fe8fd74b14045989633867b036e8b355ef57c1d78a52a63 |
|
MD5 | dcf57a0e69c0af43d6ad1beb778f308d |
|
BLAKE2b-256 | 5dd9ea3f8bc7531850e471154f2ecee60db661c6ae45f6d8fb3a48942231b3bc |