app.py

import gradio as gr
from langchain import SQLDatabase
from langchain_openai import ChatOpenAI
from langchain_experimental.sql import SQLDatabaseChain
import os
from dotenv import load_dotenv

load_dotenv()

DB_USERNAME = os.environ.get("DB_USERNAME")
DB_PASSWORD = os.environ.get("DB_PASSWORD")
DB_HOST = os.environ.get("DB_HOST")
DB_PORT = os.environ.get("DB_PORT")
DB_NAME = os.environ.get("DB_NAME")
OPENAI_API_KEY = os.environ.get("OPENAI_API_KEY")
GPT_MODEL = "gpt-3.5-turbo-0125"

if OPENAI_API_KEY is None:
    OPENAI_API_KEY = os.getenv("OPENAI_API_KEY")

db_uri = "sqlite:///carbon-emissions.db"
db = SQLDatabase.from_uri(db_uri)
llm = ChatOpenAI(api_key=OPENAI_API_KEY, model=GPT_MODEL)
db_chain = SQLDatabaseChain(llm=llm, database=db, verbose=True, top_k=3)

PROMPT = """ 
Given an input question, first create a syntactically correct Postgresql query without ```sql formatting in the query. 
Run the query then look at the results of the query.
Interpret th results and return an answer.  
The question: {question}
"""


def query(question):
    try:
        response = db_chain.run(PROMPT.format(question=question))
        return response
    except:
        return "Could not perform request. Try another one"


iface = gr.Interface(
    fn=query,
    inputs="text",
    outputs="text",
    examples=[
        ["What were the total emissions for 2022 for each category?"],
        ["What were the emissions for each year?"],
    ],
)
iface.launch()