app.py

import gradio as gr
from transformers import pipeline

# Load Wav2Vec 2.0 speech recognition model
asr = pipeline("automatic-speech-recognition", model="facebook/wav2vec2-base-960h")

# Function to process audio and return transcribed text
def transcribe(audio):
    if audio is None:
        return "No audio provided."
    return asr(audio)["text"]

# Gradio Interface
interface = gr.Interface(
    fn=transcribe,
    inputs=gr.Audio(type="filepath", label="Upload or Record Audio"),
    outputs=gr.Textbox(label="Transcribed Text"),
    title="🎤 Wav2Vec 2.0 - Speech to Text",
    description="Upload or record an audio file and get the transcribed text using Facebook's Wav2Vec 2.0."
)

interface.launch()