app.py

import gradio as gr
import os
import whisper

# Load the Whisper model
model = whisper.load_model("base")

# Function to process the uploaded audio file and perform transcription
def process_audio(upload):
    # Save the uploaded audio file
    file_path = "uploaded_audio.wav"
    with open(file_path, "wb") as f:
        f.write(upload.read())
    
    # Load the audio file and perform preprocessing
    audio = whisper.load_audio(file_path)
    audio = whisper.pad_or_trim(audio)
    mel = whisper.log_mel_spectrogram(audio).to(model.device)
    
    # Detect the spoken language
    _, probs = model.detect_language(mel)
    detected_language = max(probs, key=probs.get)
    
    # Perform transcription using Whisper ASR
    options = whisper.DecodingOptions()
    result = whisper.decode(model, mel, options)
    transcription = result.text
    
    # Delete the temporary audio file
    os.remove(file_path)
    
    return transcription

# Create a file input component for uploading the audio file
audio_input = gr.inputs.File(label="Upload Audio")

# Create a text output component for displaying the transcription
text_output = gr.outputs.Textbox(label="Transcription")

# Create a Gradio interface
gr.Interface(fn=process_audio, inputs=audio_input, outputs=text_output, title="Audio Transcription").launch()