redis-chatgpt 0.1.1

A user-friendly package designed to save chat history conversation and truncate conversation if user exceeds max token when using OpenAI chat completion.

Redis Chatgpt

Allows to save chat history of converation.

Getting started

pip install redis-chatgpt

To build a simple data store:

from redis_chatgpt.manager import RedisManager

redis_db = RedisManager(host="localhost", port=6379)

# set data
redis_key = "chat_key"
chat_memory = [
    {
      "role": "system",
      "content": "Assistant is a large language model trained by OpenAI."
    },
    {
      "role": "assistant",
      "content": "Hello! How can I assist you today?"
    },
    {
      "role": "user",
      "content": "hi"
    },
]
redis_db.set_data(redis_key, chat_memory)

# get data

chat_history = redis_db.get_data(redis_key)

# truncate history(conversation)

# You can choose whether to update the Redis database or not by specifying the 'overwrite' parameter (boolean)."
redis_db.truncate_conversation(redis_key, overwrite=True)

Use with FastAPI Swagger

cd examples/chatbot

1. Get your OpenAI API key here.
2. Copy env.example .env (set fastapi webserver workers and openai api key) cp env.example .env
3. docker-compose up -d
4. http://0.0.0.0:8012/docs