ralf-dialogue-jhuapl 0.0.4

An extension of the ralf toolkit with convenient primitives for building LLM-based dialogue agents.

ralf-dialogue

ralf-dialogue is a Python framework for quickly prototyping dialogue agents that leverage large language models (LLMs), like ChatGPT or GPT-4. It lives in the broader RALF ecosystem, and assumes the use of the ralf library, which provides primitives, constructs, and other utilities for building complex systems around large language models using a desingn paradigm that some parts of the community have begun to call composability.

A word on the broader RALF ecosystem

ralf is a Python library intended to assist developers in creating applications that involve calls to Large Language Models (LLMs). A core concept in ralf is the idea of composability, which allows chaining together LLM calls such that the output of one call can be used to form the prompt of another. ralf makes it easy to chain together both LLM-based and Python-based actions— enabling developers to construct complex information processing pipelines composed of simpler building blocks. Using LLMs in this way can lead to more capable, robust, steerable and inspectable applications.

A general framework for building applications -- including conversational AIs -- that rely on a mix of prompted large language models (LLMs) and conventional Python code and services. This can be considered a powerful form of neuro-symbolic AI. Other frameworks -- such as this one, ralf-dialogue -- build on the core ralf library, adding primatives and components for building conversational agents (or "chatbots") that can interact with users in natural language, leveraging context and accessing external knowledge stores or reasoning engines. Many of these components are still under active construction, and we're always looking for talented contributors.

Quickstart Guide

This quickstart guide is intended to get you up and running with ralf within a few minutes.

Installation

We recommend creating a Conda environment before installing the package:

conda create -n ralf python=3.10
conda activate ralf

Install from PyPI

You may install ralf-dialogue from PyPI using pip:

pip install ralf-dialogue-jhuapl

Install from Source

Alternatively, you can build the package from source. First, clone the Github repository:

git clone https://github.com/jhuapl-fomo/ralf-dialogue.git

Next, install the requirements using pip:

cd ralf-dialogue
pip install -r requirements.txt

Then, build the package using flit and install it using pip:

flit build
pip install .

Or if you would like an editable installation, you can instead use:

pip install -e .

OpenAI Configuration

ralf currently relies on language models provided by OpenAI. In order to access the models, you must store your OpenAI API key as an environment variable by executing the following in bash:

echo "export OPENAI_API_KEY='yourkey'" >> ~/.bashrc
source ~/.bashrc

Running the Demos

To test if installation was successful, try running the demo scripts:

cd demos
python simple_chat.py

This demo script will allow you to converse with a chatbot created using ralf-dialogue. Next, you can delve deeper and explore other features of the library:

python analyze_conversation.py

If the scripts execute successfully, you are good to go!