app.py

import gradio as gr
from transformers import pipeline
import requests
import json
import edge_tts
from edge_tts import VoicesManager
import asyncio
import random
import tempfile
import os
import inflect
from huggingface_hub import InferenceClient
import re
import time
from streaming_stt_nemo import Model

number_to_word = inflect.engine()

default_lang = "en"

engines = { default_lang: Model(default_lang) }

def transcribe(audio):
    lang = "en"
    model = engines[lang]
    text = model.stt_file(audio)[0]
    return text

client1 = InferenceClient("mistralai/Mixtral-8x7B-Instruct-v0.1")

system_instructions1 = "<s>[SYSTEM] Answer as WATSON, Famous Sherlock Holmes Assistant, Keep conversation short, clear, friendly and concise. The text provided is a request for a specific type of response from you, the virtual assistant. The request asks you to provide friendly responses as if You are the character WATSON From Sherlock Holmes Novels, The expectation is that I will avoid introductions and start answering the query directly, elaborating on all aspects of the request. As an AI-powered assistant, your task is to generate responses that appear as if they are created by a Human and not an AI-powered assistant. [USER]"


def model(text):
    generate_kwargs = dict(
        temperature=0.7,
        max_new_tokens=512,
        top_p=0.95,
        repetition_penalty=1,
        do_sample=True,
        seed=42,
    )
    
    formatted_prompt = system_instructions1 + text + "[WATSON]"
    stream = client1.text_generation(
        formatted_prompt, **generate_kwargs, stream=True, details=True, return_full_text=False)
    output = ""
    for response in stream:
        if not response.token.text == "</s>":
            output += response.token.text

    return output

def number_to_words(str):
    words = str.split(' ')
    result = []
    
    for word in words:
        if( any(char.isdigit() for char in word) ):
            word = number_to_word.number_to_words(word)

        result.append(word)

    final_result = ' '.join(result).replace('point', '')
    return final_result

async def respond(audio):
    user = transcribe(audio)
    reply = model(user)
    reply2 = number_to_words(reply)

    voices = await VoicesManager.create()
    voice = voices.find(Gender="Male", Locale="en-GB")
    
    communicate = edge_tts.Communicate(reply2, random.choice(voice)["Name"])
    
    with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as tmp_file:
        tmp_path = tmp_file.name
        await communicate.save(tmp_path)
    yield tmp_path



client1 = InferenceClient("mistralai/Mixtral-8x7B-Instruct-v0.1")

system_instructions1 = "[SYSTEM] Answer as Watson with a London accent male voice. You are the character and Assistant Watson, from Sherlock Holmes Novels, Keep conversation very short, clear, friendly and concise. The text provided is a request for a specific type of response from you, the virtual assistant. The request asks you to provide friendly responses as if You are the character Watson, from Sherlock Holmes Novel. The expectation is that I will avoid introductions and start answering the query directly, elaborating on all aspects of the request. As an AI-powered assistant, your task is to generate responses that appear as if they are created by a Real Person and not an AI-powered assistant. [USER]"

async def generate1(prompt):
    generate_kwargs = dict(
        temperature=0.7,
        max_new_tokens=512,
        top_p=0.95,
        repetition_penalty=1,
        do_sample=False,
    )
    formatted_prompt = system_instructions1 + prompt + "[WATSON]"
    stream = client1.text_generation(
        formatted_prompt, **generate_kwargs, stream=True, details=True, return_full_text=True)
    output = ""
    for response in stream:
        if not response.token.text == "</s>":
            output += response.token.text

    voices = await VoicesManager.create()
    voice = voices.find(Gender="Male", Locale="en-GB")
    
    communicate = edge_tts.Communicate(output, random.choice(voice)["Name"])
      
    with tempfile.NamedTemporaryFile(delete=False, suffix=".wav") as tmp_file:
        tmp_path = tmp_file.name
        await communicate.save(tmp_path)
    yield tmp_path

with gr.Blocks(css="style.css") as demo:    
    gr.Markdown(""" # <center><b>Watson 🤖 🧠 🧬</b></center>
        ### <center>An Artificial Intelligence Assistant for YOU
        ### <center>Talk to - Watson</center>
        """)
    
    with gr.Row():
        user_input = gr.Audio(label="Voice Chat", type="filepath")
        output_audio = gr.Audio(label="WATSON", type="filepath",
                        interactive=False,
                        autoplay=True,
                        elem_classes="audio")
    with gr.Row():
        translate_btn = gr.Button("Response")
        translate_btn.click(fn=respond, inputs=user_input,
                            outputs=output_audio, api_name=False)
    
    with gr.Row():
        user_input = gr.Textbox(label="Prompt", value="What's a fun science experiment we can do Watson?")
        input_text = gr.Textbox(label="Input Text", elem_id="important")
        output_audio = gr.Audio(label="WATSON", type="filepath",
                        interactive=False,
                        autoplay=True,
                        elem_classes="audio")
    with gr.Row():
        translate_btn = gr.Button("Response")
        translate_btn.click(fn=generate1, inputs=user_input,
                            outputs=output_audio, api_name="translate")  



if __name__ == "__main__":
    demo.queue(max_size=200).launch()