models.py

from huggingface_hub import InferenceClient
import gradio as gr
from transformers import pipeline
import torch
from TTS.api import TTS
import os


os.environ["COQUI_TOS_AGREED"] = "1"


class AsrBot():

    def __init__(self, model_name, lang):
        self.model_name = model_name
        self.lang = lang
    
        self.pipe = pipeline(
            task="automatic-speech-recognition",
            model=self.model_name,
            chunk_length_s=30,
            device=0,
        )
        self.pipe.model.config.forced_decoder_ids = self.pipe.tokenizer.get_decoder_prompt_ids(language=self.lang, task="transcribe")
        
    def call(self, file_path):
        if file_path is None:
            raise gr.Error("No audio file submitted! Please upload or record an audio file before submitting your request.")
    
        text = self.pipe(file_path)["text"]
        return text

class LlmBot():
    def __init__(self, model):
        self.client = InferenceClient(model)
        
    def format_prompt(self, message):
          prompt = "<s>"
          prompt += f"[INST] {message} [/INST]"
          return prompt
    
    def call(self, prompt, system_prompt, temperature=0.9, max_new_tokens=256, top_p=0.95, repetition_penalty=1.0):
        temperature = float(temperature)
        if temperature < 1e-2:
            temperature = 1e-2
        top_p = float(top_p)
    
        generate_kwargs = dict(
            temperature=temperature,
            max_new_tokens=max_new_tokens,
            top_p=top_p,
            repetition_penalty=repetition_penalty,
            do_sample=True,
            seed=42,
        )
    
        formatted_prompt = self.format_prompt(f"{system_prompt}, {prompt}")
        stream = self.client.text_generation(formatted_prompt, **generate_kwargs, details=True, return_full_text=False)
        return stream.generated_text
        
        

class TtsBot():
    def __init__(self, model):
        self.model =  TTS(model).to("cuda")
    def call(self, text):
        tts = TTS("tts_models/multilingual/multi-dataset/xtts_v2").to("cuda")
        tts.tts_to_file(text=text, speaker_wav="./titty-sprinkles-101soundboards.mp3", language="en", file_path="./output.wav")
        return "./output.wav"


asr_bot = AsrBot("openai/whisper-small", "en")
llm_bot = LlmBot("mistralai/Mixtral-8x7B-Instruct-v0.1")
tts_bot = TtsBot("tts_models/multilingual/multi-dataset/xtts_v2")