Create app.py

app.py

@@ -0,0 +1,60 @@
+import os
+from transformers import pipeline
+import gradio as gr
+import torch
+import pytube as pt
+
+checkpoint = "GGmorello/whisper-small-it"
+device = 0 if torch.cuda.is_available() else "cpu"
+print(device)
+
+pipe = pipeline(task = "automatic-speech-recognition", model = checkpoint,chunk_length_s=30,device = device)
+
+def transcribe(audio):
+    text = pipe(audio)["text"]
+    return text
+
+def transcribe_url(yt_url):
+  yt = pt.YouTube(yt_url)
+  stream = yt.streams.filter(only_audio=True)[0]
+  stream.download(filename = "audio.mp3")
+  text = pipe("audio.mp3")["text"]
+  return text
+
+
+demo = gr.Blocks()
+
+microphone_interface = gr.Interface(
+    fn=transcribe,
+    inputs = gr.Audio(sources="microphone", type="filepath"),
+    outputs="text",
+    title="Whisper Small Italian Finetuned raw microphone audio",
+    description="Realtime demo for Italian speech recognition using a fine-tuned Whisper small model."
+
+)
+
+
+file_interface = gr.Interface(
+    fn=transcribe,
+    inputs = gr.Audio(sources="upload", type="filepath"),
+    outputs="text",
+    title="Whisper Small Italian Finetuned for audio file.",
+    description="Realtime demo for Italian speech recognition using a fine-tuned Whisper small model."
+
+)
+
+
+url_interface = gr.Interface(
+    fn = transcribe_url,
+    inputs = gr.Textbox(lines=1, placeholder="Paste the URL to a YouTube video here", label="YouTube URL"),
+    outputs = "text",
+    title = "Whisper Small Italian Finetuned for URL transcription",
+    description = "Realtime demo for Italian speech recognition using a fine-tuned Whisper small model."
+
+
+)
+
+with demo:
+    gr.TabbedInterface([microphone_interface,file_interface, url_interface], ["Transcribe Audio", "Transcribe File" , "Transcribe YouTube"])
+
+demo.launch(share=True)