PromptSource: An Integrated Development Environment and Repository for Natural Language Prompts

TL;DR

PromptSource is a comprehensive system that facilitates the creation, sharing, and refinement of natural language prompts for NLP tasks, supporting collaborative development and a large repository of prompts.

Contribution

It introduces a templating language, an interactive interface, and community guidelines, enabling efficient prompt development and sharing in NLP.

Findings

Over 2,000 prompts for 170 datasets available

Supports collaborative prompt creation and refinement

Provides a user-friendly interface for prompt iteration

Abstract

PromptSource is a system for creating, sharing, and using natural language prompts. Prompts are functions that map an example from a dataset to a natural language input and target output. Using prompts to train and query language models is an emerging area in NLP that requires new tools that let users develop and refine these prompts collaboratively. PromptSource addresses the emergent challenges in this new setting with (1) a templating language for defining data-linked prompts, (2) an interface that lets users quickly iterate on prompt development by observing outputs of their prompts on many examples, and (3) a community-driven set of guidelines for contributing new prompts to a common pool. Over 2,000 prompts for roughly 170 datasets are already available in PromptSource. PromptSource is available at https://github.com/bigscience-workshop/promptsource.