ChatGPT Clone Project

언어 선택 / Language Selection:

• 🇰🇷 한국어 (Korean)
• 🇺🇸 English

Technology Stack and Features

• ⚡ FastAPI for the Python backend API.
  • 🧰 SQLAlchemy for the Python SQL database interactions (ORM).
  • 🔍 Pydantic used by FastAPI, for the data validation and settings management.
  • 💾 PostgreSQL as the SQL database.
  • 📁 Adminer as the Database Management System
  • 🤖 Ollama as the local hosting server for LLM(EEVE-Korean)
  • ⛓️ LangChain to build LLM chat bot
• 🚀 Svelte for the frontend
• 🐋 Docker Compose for development and production.
• 🔒 Secure password hashing by default.
• 🔑 JWT (JSON Web Token) authentication.
• 📫 Email based password recovery.
• ✅ Tests with Pytest.
• 📞 Traefik as a reverse proxy / load balancer.
• 🚢 Deployment instructions using Docker Compose, including how to set up a frontend Traefik proxy to handle automatic HTTPS certificates.

Dashboard Login

Dashboard password recovery

Dashboard Chat

How to use it

Infra

I developed this project using the GPU cloud service(paperspace)

• OS: Ubuntu 22.04
• GPU: Quadro RTX4000(8192MiB)
  • It could be difficult to use RAG because of out of memory in GPU.
• CUDA version: 12.6

.env file setting

please make the .env file in the root folder

PROJECT_NAME="ChatGPT Clone Project"
STACK_NAME="ChatGPT-Clone-Project"
DOMAIN=localhost

# backend url
VITE_SERVER_URL=http://127.0.0.1:8000
VITE_EEVE_SERVER_URL=http://127.0.0.1:9000

# frontend url
BACKEND_CORS_ORIGINS="http://localhost,http://localhost:5173,http://127.0.0.1:5173,https://localhost,https://localhost:5173,https://127.0.0.1:5173"
DOMAIN_PORT="5173"

USE_HASH_ROUTER = "True"
ACCESS_TOKEN_EXPIRE_MINUTES = 60

# secret key and algorithm for auth
SECRET_KEY =
ALGORITHM =

SMTP_HOST = "smtp.gmail.com"
SMTP_PORT = 587
SMTP_USERNAME =
SMTP_PASSWORD =
EMAILS_FROM_EMAIL = "[email protected]"
EMAILS_FROM_NAME = "ChatGPT Clone Project Information"

# Postgres
## Dev(or Prod) DB
POSTGRES_SERVER=localhost
POSTGRES_PORT=5432
POSTGRES_DB=app
POSTGRES_USER=postgres
POSTGRES_PASSWORD=changethis
## Test DB
TEST_POSTGRES_SERVER=localhost
TEST_POSTGRES_PORT=5432
TEST_POSTGRES_DB=test
TEST_POSTGRES_USER=postgres
TEST_POSTGRES_PASSWORD=changethis

# langsmith
LANGCHAIN_TRACING_V2 = 'true'
LANGCHAIN_API_KEY =

• PROJECT_NAME: The name of the project
• STACK_NAME: The name of the stack used for Docker Compose labels and project name (no spaces, no periods) (in .env).
• SECRET_KEY: The secret key for the project, used for security, stored in .env.
• SMTP_USERNAME: The SMTP server user to send emails.
• SMTP_PASSWORD: The SMTP server password to send emails.
• LANGCHAIN_TRACING_V2: Where to use Langsmith to tracing AI chat bot's response in detail.
• LANGCHAIN_API_KEY: API Key to tracing AI chat bot's response in Langsmith.

Execute Containers using docker compsoe

sudo docker-compose up

it will make the containeres

• reverse proxy(Traefik)
• Database(PostgreSQL)
• backend(FastAPI)
• frontend(Svelte)
• Database Management System(Adminer)

for example)

[+] Building 0.0s (0/0)                                                                                                                                                               docker:default
[+] Running 5/0
 ✔ Container chatgpt-clone-proxy-1    Created                                                                                                                                                   0.0s
 ✔ Container chatgpt-clone-db-1       Created                                                                                                                                                   0.0s
 ✔ Container backend                  Created                                                                                                                                                   0.0s
 ✔ Container frontend                 Created                                                                                                                                                   0.0s
 ✔ Container chatgpt-clone-adminer-1  Created                                                                                                                                                   0.0s
Attaching to backend, chatgpt-clone-adminer-1, chatgpt-clone-db-1, chatgpt-clone-proxy-1, frontend

Running Containers in a CPU-Only Environment

If you are running in a CPU-only environment then you cannot use the EEVE-Korean model, you can start all other containers except for the eeve container.

To do this, use the following command:

docker compose up db adminer backend frontend proxy

or you can use this following command:

sudo docker compose up --scale eeve=0

Note: The EEVE-Korean model will not be available for interaction.

Excute Containers using docker compose in test environment

To run containers in the test environment, use the following command:

sudo docker-compose -f docker-compose.yaml -f docker-compose.override.yaml -f docker-compose.test.yaml up

Running this command will start a test database (test DB) that is isolated from the development (dev) and production (prod) databases.

When running tests in the backend or eeve, all test-related data will be stored in the test DB. To ensure data separation during testing, it is strongly recommended to use the test DB.

Backend Development

Backend docs: backend/readme.md

Development

General development docs: development.md.

This includes using Docker Compose, pre-commit, .env configurations, etc.

Acknowledgements

This repository is built upon full-stack-fastapi-template. If you want to use FastAPI, that will be good reference or starting point.

@online{full-stack-fastapi-template,
  author    = {fastapi},
  title     = {full-stack-fastapi-template},
  url       = {https://github.com/fastapi/full-stack-fastapi-template},
  year      = {2024},
}