Open-Source Documentation Assistant
DocsGPT is a cutting-edge open-source solution that streamlines the process of finding information in project documentation. With its integration of the powerful GPT models, developers can easily ask questions about a project and receive accurate answers.
Say goodbye to time-consuming manual searches, and let DocsGPT help you quickly find the information you need. Try it out and see how it revolutionizes your project documentation experience. Contribute to its development and be a part of the future of AI-powered assistance.
We're eager to provide personalized assistance when deploying your DocsGPT to a live environment.
You can find our roadmap here. Please don't hesitate to contribute or create issues, it helps us improve DocsGPT!
Name | Base Model | Requirements (or similar) |
---|---|---|
Docsgpt-7b-falcon | Falcon-7b | 1xA10G gpu |
Docsgpt-14b | llama-2-14b | 2xA10 gpu's |
Docsgpt-40b-falcon | falcon-40b | 8xA10G gpu's |
If you don't have enough resources to run it, you can use bitsnbytes to quantize.
-
🔍🔥 Live preview
-
📚😎 Guides
-
👩💻👨💻 Interested in contributing?
-
🏠🔐 How to host it locally (so all data will stay on-premises)
-
Application - Flask app (main application).
-
Extensions - Chrome extension.
-
Scripts - Script that creates similarity search index and stores for other libraries.
-
Frontend - Frontend uses Vite and React.
Note: Make sure you have Docker installed
On Mac OS or Linux, write:
./setup.sh
It will install all the dependencies and allow you to download the local model or use OpenAI.
Otherwise, refer to this Guide:
-
Download and open this repository with
git clone https://github.com/arc53/DocsGPT.git
-
Create a
.env
file in your root directory and set the env variableOPENAI_API_KEY
with your OpenAI API key andVITE_API_STREAMING
to true or false, depending on if you want streaming answers or not. It should look like this inside:API_KEY=Yourkey VITE_API_STREAMING=true
See optional environment variables in the /.env-template and /application/.env_sample files.
-
Navigate to http://localhost:5173/.
To stop, just run Ctrl + C
.
For development, only two containers are used from docker-compose.yaml (by deleting all services except for Redis and Mongo). See file docker-compose-dev.yaml.
Run
docker compose -f docker-compose-dev.yaml build
docker compose -f docker-compose-dev.yaml up -d
Make sure you have Python 3.10 or 3.11 installed.
- Export required environment variables or prepare a
.env
file in the/application
folder:- Copy .env_sample and create
.env
with your OpenAI API token for theAPI_KEY
andEMBEDDINGS_KEY
fields.
- Copy .env_sample and create
(check out application/core/settings.py
if you want to see more config options.)
- (optional) Create a Python virtual environment: You can follow the Python official documentation for virtual environments .
a) On Mac OS and Linux
python -m venv venv
. venv/bin/activate
b) On Windows
python -m venv venv
venv/Scripts/activate
- Change to the
application/
subdir by the commandcd application/
and install dependencies for the backend:
pip install -r requirements.txt
- Run the app using
flask run --host=0.0.0.0 --port=7091
. - Start worker with
celery -A application.app.celery worker -l INFO
.
Make sure you have Node version 16 or higher.
- Navigate to the /frontend folder.
- Install required packages
husky
andvite
(ignore if installed).
npm install husky -g
npm install vite -g
- Install dependencies by running
npm install --include=dev
. - Run the app using
npm run dev
.
Please refer to the CONTRIBUTING.md file for information about how to get involved. We welcome issues, questions, and pull requests.
We as members, contributors, and leaders, pledge to make participation in our community a harassment-free experience for everyone, regardless of age, body size, visible or invisible disability, ethnicity, sex characteristics, gender identity and expression, level of experience, education, socio-economic status, nationality, personal appearance, race, religion, or sexual identity and orientation. Please refer to the CODE_OF_CONDUCT.md file for more information about contributing.
The source code license is MIT, as described in the LICENSE file.
Built with 🦜️🔗 LangChain