GPT4All

How to configure and use it?

Pre-Requisite(s)

• model (optional) - Set which model you want to use. Defaults to orca-mini-3b.ggmlv3.q4_0

Running in a Colab/Kaggle/Python scripts(s)

from genai_stack.model import Gpt4AllModel

llm = Gpt4AllModel.from_kwargs()
model_response = llm.predict("How many countries are there in the world?")
print(model_response["result"])

Running the model in a webserver

If you want to run the model in a webserver and interact with it with HTTP requests, the model provides a way to run it.

1. As a Python script

from fastapi.responses import JSONResponse
from genai_stack.model import Gpt4AllModel

llm = Gpt4AllModel.from_kwargs()
llm.run_http_server(response_class=JSONResponse)

A server should start as below

INFO:     Started server process [137717]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://127.0.0.1:8082 (Press CTRL+C to quit)

Make HTTP requests. URL - http://localhost:8082/predict/

import requests
response = requests.post("http://localhost:8082/predict/", data="How many countries are there in the world?")
print(response.text)

1. As a CLI

Create a model.json file with the following contents:

{
    "model": {
        "name": "gpt4all",
        "fields": {
            "model": "ggml-gpt4all-j-v1.3-groovy"
        }
    }
}

Run the below command:

genai-stack start --config_file model.json

 ██████╗ ███████╗███╗   ██╗ █████╗ ██╗    ███████╗████████╗ █████╗  ██████╗██╗  ██╗    
██╔════╝ ██╔════╝████╗  ██║██╔══██╗██║    ██╔════╝╚══██╔══╝██╔══██╗██╔════╝██║ ██╔╝    
██║  ███╗█████╗  ██╔██╗ ██║███████║██║    ███████╗   ██║   ███████║██║     █████╔╝     
██║   ██║██╔══╝  ██║╚██╗██║██╔══██║██║    ╚════██║   ██║   ██╔══██║██║     ██╔═██╗     
╚██████╔╝███████╗██║ ╚████║██║  ██║██║    ███████║   ██║   ██║  ██║╚██████╗██║  ██╗    
 ╚═════╝ ╚══════╝╚═╝  ╚═══╝╚═╝  ╚═╝╚═╝    ╚══════╝   ╚═╝   ╚═╝  ╚═╝ ╚═════╝╚═╝  ╚═╝
INFO:     Started server process [641734]
INFO:     Waiting for application startup.
INFO:     Application startup complete.
INFO:     Uvicorn running on http://127.0.0.1:8082 (Press CTRL+C to quit)