💵 What is it?
Index Compression Methods (INCOME) repository helps you easily train and evaluate memory-compressed binary retrievers on any custom dataset. We provide recent state-of-the-art techniques for training and unsupervised (without requiring custom training data) for domain-adaptation of NLP-based binary retrieval models across any dataset.
For more information, checkout our publication:
- Domain Adaptation for Memory-Efficient Dense Retrieval (Arxiv preprint)
💵 Installation
One can either install income via pip
pip install income
or via source using git clone
$ git clone https://github.com/Nthakur20/income.git
$ cd income
$ pip install -e .
💵 Models Supported
Uploaded Public Models
💵 Quick Example
💵 Why should we do domain adaptation?
💵 Inference
💵 Training
💵 Citing & Authors
If you find this repository helpful, feel free to cite our recent publication: Domain Adaptation for Memory-Efficient Dense Retrieval:
@article{thakur2022domain,
title={Domain Adaptation for Memory-Efficient Dense Retrieval},
author={Thakur, Nandan and Reimers, Nils and Lin, Jimmy},
journal={arXiv preprint arXiv:2205.11498},
year={2022},
url={https://arxiv.org/abs/2205.11498/}
}
The main contributors of this repository are:
- Nandan Thakur, Personal Website: nandan-thakur.com
Contact person: Nandan Thakur, [email protected]
Don’t hesitate to send us an e-mail or report an issue, if something is broken (and it shouldn’t be) or if you have further questions.
This repository contains experimental software and is published for the sole purpose of giving additional background details on the respective publication.
