Skip to content

installation.md

Latest commit

 

History

History
106 lines (52 loc) · 4.64 KB

installation.md

File metadata and controls

106 lines (52 loc) · 4.64 KB

Installation

Components

The system comprises seven main components, here are their relationships and the technologies they are built with:

ChatFAQ Components

  • Back-end manages the communication between all the components. It also houses the database for storing all the data related to the chatbots, datasets, models, etc...

  • SDK launches a Remote Procedure Call (RPC) server to execute transitions and events from the posted FSM definitions.

  • Widget is a JS browser client application from which the user interacts with the bot.

  • Admin is a JS browser client application to manage the chatbots, knowledge bases, retrievers, models, configs, etc...

  • Ray workers are used to run distributed inference on the models.

  • Channel layer (Redis) is used to communicate through WebSockets between the back-end and the SDK, admin and widget.

  • Relational Database (PostgreSQL) is used to store all the data related to the chatbots, knowledge bases, retrievers, models, configs, etc...

Docker

Before running the services, ensure you have Docker and Docker Compose installed on your system. You can install them using the following instructions:

We prepared you a docker-compose.yaml that set up all the services for you. You can find it on the root of the repository.

Configuration Instructions

But first you add to your hosts file (usually under /etc/hosts) the following lines in order to share the .env files values between a local deployment and a docker deployment:

127.0.0.1 postgres 127.0.0.1 back 127.0.0.1 redis

Then you need to create the corresponding .env files for each service. You can see an example of those on:

Now you can run the migrations and the fixtures for the initial database data:

docker compose -f docker-compose.yaml -f docker-compose.vars.yaml run back poetry run ./manage.py migrate
docker compose -f docker-compose.yaml -f docker-compose.vars.yaml run back poetry run make apply_fixtures

Create a superuser on the backend (making sure you answer 'yes' to the question 'Belongs to the RPC group?')

docker compose -f docker-compose.yaml -f docker-compose.vars.yaml run back poetry run ./manage.py createsuperuser

Generate a ChatFAQ Token with the user and password you just created:

docker compose -f docker-compose.yaml -f docker-compose.vars.yaml run back poetry run ./manage.py createtoken <USER>

Which will respond something as such:

Token for user <USER> created: <TOKEN>

Add it to your sdk/.env file:

CHATFAQ_TOKEN=<TOKEN>

and the back/.env file:

BACKEND_TOKEN=<TOKEN>

The configuration we provided for the LLM uses an OpenAI model. To use it, you need to add your OpenAI API key to the back/.env file:

OPENAI_API_KEY=<API_KEY>

Don't worry, our solution supports any other LLM model. It also supports deploying your own local model on a VLLM server. However, for simplicity and because OpenAI models are the most popular, we are using an OpenAI model as the default.

Running the Services

To start all the services, run the following command:

docker compose -f docker-compose.yaml -f docker-compose.vars.yaml up

Congratulations! You now have a running ChatFAQ instance.

Accessing the Chatbot

You can interact with the chatbot by navigating to:

http://localhost:3003/demo/

To manage the chatbot and view the model configuration we provided, go to:

http://localhost:3000

Deeper into ChatFAQ

If you want to use your own knowledge base, you can check the Knowledge Base Configuration documentation.

If you want to learn how to use the SDK, so you can create your own chatbot behavior, you can check the SDK documentation.