app.py

from multilingual_translation import text_to_text_generation
from utils import lang_ids, data_scraping
import whisper
import gradio as gr

lang_list = list(lang_ids.keys())
model_list = data_scraping()
model = whisper.load_model("small")

def transcribe(audio):
    
    #time.sleep(3)
    # load audio and pad/trim it to fit 30 seconds
    audio = whisper.load_audio(audio)
    audio = whisper.pad_or_trim(audio)

    # make log-Mel spectrogram and move to the same device as the model
    mel = whisper.log_mel_spectrogram(audio).to(model.device)

    # detect the spoken language
    _, probs = model.detect_language(mel)
    print(f"Detected language: {max(probs, key=probs.get)}")

    # decode the audio
    options = whisper.DecodingOptions(fp16 = False)
    result = whisper.decode(model, mel, options)

    finalResult = text_to_text_generation(prompt='return.text', model_id='facebook/m2m100_418M', device='cpu',target_lang='English')
    return finalResult

# api endpoint to return the transcription in EN as a json response

# @app.route('/transcribe', methods=['POST'])
# def transcribe_api():
#     if request.method == 'POST':
#         audio = request.files['audio']
#         audio = audio.read()
#         audio = io.BytesIO(audio)
#         audio = whisper.load_audio(audio)
#         audio = whisper.pad_or_trim(audio)
#         mel = whisper.log_mel_spectrogram(audio).to(model.device)
#         _, probs = model.detect_language(mel)
#         print(f"Detected language: {max(probs, key=probs.get)}")
#         options = whisper.DecodingOptions(fp16 = False)
#         result = whisper.decode(model, mel, options)
#         return jsonify(result)


    
    
 
gr.Interface(
    title = 'OpenAI Whisper ASR Gradio Web UI', 
    fn=transcribe, 
    inputs=[
        gr.inputs.Audio(source="microphone", type="filepath")
    ],
    outputs=[
        "textbox"
    ],
    live=True).launch(debug=True, enable_queue=True)

# output = gr.outputs.Textbox(label="Output Text")