app.py

import whisper
model = whisper.load_model("base")

import deepl
deepl_auth_key = os.environ["Deepl_API"]

def translate(text, target_lang):
  translator = deepl.Translator(deepl_auth_key)
  translated_text = translator.translate_text(text, target_lang=target_lang)
  return translated_text

def transcribe(audio):

    # 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)}")
    detect_lang = max(probs, key=probs.get)

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


    translated_text = translate(result.text, "JA")
    return translated_text

import gradio as gr

title = 'Video Translator'

inputs = gr.Video()
outputs = gr.Text()
interface = gr.Interface(title=title, fn=transcribe, inputs=inputs, outputs=outputs)
interface.launch(debug=True)