Update app.py

app.py

@@ -1,11 +1,9 @@
 import torch
-
 import gradio as gr
 import spaces
 import yt_dlp as youtube_dl
 from transformers import pipeline
 from transformers.pipelines.audio_utils import ffmpeg_read
-
 import tempfile
 import os
 
@@ -29,8 +27,7 @@ def transcribe(inputs, task):
         raise gr.Error("No audio file submitted! Please upload or record an audio file before submitting your request.")
 
     text = pipe(inputs, batch_size=BATCH_SIZE, generate_kwargs={"task": task}, return_timestamps=True)["text"]
-    return  text
-
+    return text
 
 def _return_yt_html_embed(yt_url):
     video_id = yt_url.split("?v=")[-1]
@@ -63,7 +60,14 @@ def download_yt_audio(yt_url, filename):
         file_length_hms = time.strftime("%HH:%MM:%SS", time.gmtime(file_length_s))
         raise gr.Error(f"Maximum YouTube length is {yt_length_limit_hms}, got {file_length_hms} YouTube video.")
     
-    ydl_opts = {"outtmpl": filename, "format": "worstvideo[ext=mp4]+bestaudio[ext=m4a]/best[ext=mp4]/best"}
+    ydl_opts = {
+        "outtmpl": filename,
+        "format": "bestaudio/best",
+        "postprocessors": [{
+            "key": "FFmpegExtractAudio",
+            "preferredcodec": "mp3",
+        }]
+    }
     
     with youtube_dl.YoutubeDL(ydl_opts) as ydl:
         try:
@@ -71,12 +75,11 @@ def download_yt_audio(yt_url, filename):
         except youtube_dl.utils.ExtractorError as err:
             raise gr.Error(str(err))
 
-
 def yt_transcribe(yt_url, task, max_filesize=75.0):
     html_embed_str = _return_yt_html_embed(yt_url)
 
     with tempfile.TemporaryDirectory() as tmpdirname:
-        filepath = os.path.join(tmpdirname, "video.mp4")
+        filepath = os.path.join(tmpdirname, "audio.mp3")
         download_yt_audio(yt_url, filepath)
         with open(filepath, "rb") as f:
             inputs = f.read()
@@ -86,8 +89,7 @@ def yt_transcribe(yt_url, task, max_filesize=75.0):
 
     text = pipe(inputs, batch_size=BATCH_SIZE, generate_kwargs={"task": task}, return_timestamps=True)["text"]
 
-    return html_embed_str, text
-
+    return html_embed_str, text, filepath
 
 demo = gr.Blocks()
 
@@ -131,7 +133,11 @@ yt_transcribe = gr.Interface(
         gr.Textbox(lines=1, placeholder="Paste the URL to a YouTube video here", label="YouTube URL"),
         gr.Radio(["transcribe", "translate"], label="Task", value="transcribe")
     ],
-    outputs=["html", "text"],
+    outputs=[
+        "html",
+        "text",
+        gr.Audio(label="Extracted Audio")
+    ],
     theme="huggingface",
     title="Whisper Large V3: Transcribe YouTube",
     description=(
@@ -145,5 +151,4 @@ yt_transcribe = gr.Interface(
 with demo:
     gr.TabbedInterface([mf_transcribe, file_transcribe, yt_transcribe], ["Microphone", "Audio file", "YouTube"])
 
-demo.launch()
-
+demo.launch()