Continual-Learning-for-LLMs

Continual Learning for Large Language models

Setup

Using Docker

1. Setup .env file use example from. env.example. Note: donot add db_host manually.

db_username = 'postgres'
db_password = 'postgres'
db_name = 'memory_manager'
db_host = db
db_port = '5432'
conversation_table_name = 'chat_history'
OPENAI_API_KEY = 'sk-<your-openai-api-key>'

2. docker compose up

Without Docker

Postgres Setup

1. Download postgres - https://www.postgresql.org/download/
2. Create new database for the project

Python

1. setup .env file

db_username = 'postgres'
db_password = 'postgres'
db_name = 'memory_manager'
db_host = '127.0.0.1'
db_port = '5432'
conversation_table_name = 'chat_history'
OPENAI_API_KEY = 'sk-<your-openai-api-key>'

2. Setup enviornment

conda create -n cll python=3.11 

conda activate cll

3. Go to backend folder

cd continual_learning/backend

4. Install the packages from requirements.txt

pip install -r requirements.txt

5. Run server

python backend.py

Frontend

1. Go to frontend folder

cd continual_learning/frontend

2. Install packages

npm i

3. Run server

npm run dev