A framework for composing retrieval models and language models into powerful pipelines that tackle knowledge-intensive tasks.
Demonstrate-Search-Predict: Composing retrieval and language models for knowledge-intensive NLP
pip install dsp-ml
Our intro notebook provides examples of five "multi-hop" question answering programs of increasing complexity written in DSP.
You can open the notebook in Google Colab. You don't even need an API key to get started with it.
Once you go through the notebook, you'll be ready to create your own DSP pipelines!
If you use DSP in a research paper, please cite our work as follows:
@article{khattab2022demonstrate,
title={Demonstrate-Search-Predict: Composing Retrieval and Language Models for Knowledge-Intensive {NLP}},
author={Khattab, Omar and Santhanam, Keshav and Li, Xiang Lisa and Hall, David and Liang, Percy and Potts, Christopher and Zaharia, Matei},
journal={arXiv preprint arXiv:2212.14024},
year={2022}
}