TalktoChatGPT1

import gradio as gr
import openai
import pyttsx3

openai.api_key = "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"

# Global variable to hold the chat history, initialise with system role
conversation = [
        {"role": "system", "content": "You are an intelligent professor."}
        ]

# Add your construct_index function here
def construct_index(directory_path):
    max_input_size = 4096
    num_outputs = 512
    max_chunk_overlap = 20
    chunk_size_limit = 2048

    prompt_helper = PromptHelper(max_input_size, num_outputs, max_chunk_overlap, chunk_size_limit=chunk_size_limit)

    llm_predictor = LLMPredictor(llm=ChatOpenAI(temperature=0.7, model_name="gpt-3.5-turbo", max_tokens=num_outputs))

    documents = SimpleDirectoryReader(directory_path).load_data()

    index = GPTSimpleVectorIndex(documents, llm_predictor=llm_predictor, prompt_helper=prompt_helper)

    index.save_to_disk('index.json')

    return index

# Construct the index here
directory_path = "path_to_your_directory"  # replace with your directory path
index = construct_index(directory_path)

# transcribe function to record the audio input
def transcribe(audio):
    print(audio)

    # Whisper API
    audio_file = open(audio, "rb")
    transcript = openai.Audio.transcribe("whisper-1", audio_file)

    print(transcript)

    # append user's input to conversation
    conversation.append({"role": "user", "content": transcript["text"]})

    response = openai.ChatCompletion.create(
    model="gpt-3.5-turbo",
    messages=conversation
    )
    
    print(response)

    # system_message is the response from ChatGPT API
    system_message = response["choices"][0]["message"]["content"]

    # append ChatGPT response (assistant role) back to conversation
    conversation.append({"role": "assistant", "content": system_message})

    # Query the index
    index_answer = index.query(transcript["text"])

    # Text to speech
    engine = pyttsx3.init()
    engine.setProperty("rate", 150)
    engine.setProperty("voice", "english-us")
    engine.save_to_file(system_message, "response.mp3")
    engine.runAndWait()

    # return response as text and audio
    return transcript["text"], system_message, index_answer, "response.mp3"

# Gradio output
bot = gr.Interface(
    fn=transcribe,
    inputs=gr.Audio(source="microphone", type="filepath"),
    outputs=[gr.outputs.Textbox(label="Transcribed Text"), gr.outputs.Textbox(label="API Answer"), gr.outputs.Textbox(label="Index Answer"), "audio"],
)
bot.launch()