app.py

import soundfile as sf
import torch
from transformers import Wav2Vec2ForCTC, Wav2Vec2Processor
import gradio as gr
import sox
import subprocess
from fuzzywuzzy import fuzz
from data import get_data


DATASET = get_data()

def read_file_and_process(wav_file):
    filename = wav_file.split('.')[0]
    filename_16k = filename + "16k.wav"
    resampler(wav_file, filename_16k)
    speech, _ = sf.read(filename_16k)
    inputs = processor(speech, sampling_rate=16_000, return_tensors="pt", padding=True)
    
    return inputs


def resampler(input_file_path, output_file_path):
    command = (
        f"ffmpeg -hide_banner -loglevel panic -i {input_file_path} -ar 16000 -ac 1 -bits_per_raw_sample 16 -vn "
        f"{output_file_path}"
    )
    subprocess.call(command, shell=True)


def parse_transcription(logits):
    predicted_ids = torch.argmax(logits, dim=-1)
    transcription = processor.decode(predicted_ids[0], skip_special_tokens=True)
    return transcription


def parse(wav_file):
    input_values = read_file_and_process(wav_file)
    with torch.no_grad():
        logits = model(**input_values).logits
    user_question = parse_transcription(logits)
    return user_question


# Function to retrieve an answer based on a question (using fuzzy matching)
def get_answer(wav_file=None):
    
    input_values = read_file_and_process(wav_file)
    
    with torch.no_grad():
        logits = model(**input_values).logits
    user_question = parse_transcription(logits)
    
    highest_score = 0
    best_answer = None

    for item in DATASET:
        similarity_score = fuzz.token_set_ratio(user_question, item["question"])
        if similarity_score > highest_score:
            highest_score = similarity_score
            best_answer = item["answer"]

    if highest_score >= 80:  # Adjust the similarity threshold as needed
        return best_answer
    else:
        return "I don't have an answer to that question."


model_id = "jonatasgrosman/wav2vec2-large-xlsr-53-persian"
processor = Wav2Vec2Processor.from_pretrained(model_id)
model = Wav2Vec2ForCTC.from_pretrained(model_id)

input_ = [
        gr.Audio(source="microphone",
                  type="filepath",
                  label="لطفا دکمه ضبط صدا را بزنید و شروع به صحبت کنید و بعذ از اتمام صحبت دوباره دکمه ضبط را فشار دهید.",
                  show_download_button=True,
                  show_edit_button=True,
                 ), 
        # gr.Textbox(label="سوال خود را بنویسید.",
        #            lines=3,
        #            text_align="right",
        #            show_label=True,)
         ]

txtbox = gr.Textbox(
            label="پاسخ شما: ",
            lines=5,
            text_align="right",
            show_label=True,
            show_copy_button=True,
        )

title = "Speech-to-Text (persian)"
description = "، توجه داشته باشید که هرچه گفتار شما شمرده تر باشد خروجی با کیفیت تری دارید.روی دکمه ضبط صدا کلیک کنید و سپس دسترسی مرورگر خود را به میکروفون دستگاه بدهید، سپس شروع به صحبت کنید و برای اتمام ضبط دوباره روی دکمه کلیک کنید"
article = "<p style='text-align: center'><a href='https://github.com/nimaprgrmr'>Large-Scale Self- and Semi-Supervised Learning for Speech Translation</a></p>"

demo = gr.Interface(fn=get_answer, inputs = input_,  outputs=txtbox, title=title, description=description, article = article,
             streaming=True, interactive=True,
             analytics_enabled=False, show_tips=False, enable_queue=True)
demo.launch(share=True)