app.py

import os
import openai
import gradio as gr
from transformers import AutoModelForCausalLM, AutoTokenizer
import torch

openai.api_key = os.getenv("OPENAI_API_KEY")

    def predict(input, history=[]):
        
        new_user_input_ids = input
        response = openai.Completion.create(
        model="davinci:ft-placeholder-2022-12-10-04-13-26",
        prompt=input
        temperature=0.13,
        max_tokens=310,
        top_p=1,
        frequency_penalty=0.36,
        presence_penalty=1.25
        )

    # generate a response 
        history = model.generate(bot_input_ids, max_length=1000, pad_token_id=tokenizer.eos_token_id).tolist()
    
    # convert the tokens to text, and then split the responses into lines
        response = tokenizer.decode(history[0]).split("<|endoftext|>")
        response = [(response[i], response[i+1]) for i in range(0, len(response)-1, 2)]  # convert to tuples of list
        return response, history




    new_user_input_ids = tokenizer.encode(input + tokenizer.eos_token, return_tensors='pt')
    
    # append the new user input tokens to the chat history
    bot_input_ids = torch.cat([torch.LongTensor(history), new_user_input_ids], dim=-1)

    # generate a response 
    response = openai.Completion.create(
    model="text-davinci-003",
    #model="davinci:ft-placeholder:ai-dhd-2022-12-07-10-09-37",
    prompt= input,
    temperature=0.09,
    max_tokens=608,
    top_p=1,
    frequency_penalty=0,
    presence_penalty=0).tolist()

gr.Interface(fn=predict,
             inputs=["text", "state"],
             outputs=["chatbot", "state"]).launch()