Update app.py

app.py

@@ -1,21 +1,47 @@
-import gradio as gr
 from transformers import pipeline
+import gradio as gr
+import os
+import deepl
+import openai
+from pytube import YouTube
 
+TARGET_LANG = "EN-GB"
+deepl_key = os.environ.get('DEEPL_KEY')
+
+translator = deepl.Translator(deepl_key)
 pipe = pipeline("automatic-speech-recognition", model="FredBonux/whisper-small-it")
 
 def transcribe(audio):
-    text = pipe(audio)["text"]
-    return text
-
-# demo = gr.Interface.from_pipeline(pipe)
-# demo.launch()
-
-iface = gr.Interface(
-    fn=transcribe,
-    inputs=gr.Audio(sources=["microphone"], type="filepath"),
-    outputs="text",
-    title="Whisper Small Italian",
-    description="Realtime demo for Italian speech recognition using a fine-tuned Whisper small model.",
+    ita = pipe(audio)["text"]
+    eng = translator.translate_text(ita, target_lang=TARGET_LANG).text
+    print(f"{ita} -> {text_en}")
+    return ira, eng
+
+def transcribe_url(url):
+    youtube = YouTube(str(url))
+    audio = youtube.streams.filter(only_audio=True).first().download('yt_video')
+    text_it = pipe(audio)["text"]
+    text_en = translator.translate_text(text_sv, target_lang=TARGET_LANG).text
+    return text_sv, text_en
+
+url_demo = gr.Interface(
+    fn=transcribe_url, 
+    inputs="text", 
+    outputs=[gr.Textbox(label="Transcribed text"),
+             gr.Textbox(label="English translation")],
+    title="Italian video to english text",
+    description="Transcribing italian video to text and translating it to english!",
+)
+
+voice_demo = gr.Interface(
+    fn=transcribe, 
+    inputs=gr.Audio(source="microphone", type="filepath"), 
+    outputs=[gr.Textbox(label="Transcribed text"),
+             gr.Textbox(label="English translation")],
+    title="Italian recorded speech to english text",
+    description="Transcribing italian speech to text and translating it to english!",
 )
 
-iface.launch()
+app = gr.TabbedInterface([url_demo, voice_demo], ["Video to English Text", "Audio to English Text"])
+
+app.launch()