Update app.py

app.py

@@ -26,14 +26,39 @@ pipe = pipeline(
 
 @spaces.GPU
 def transcribe(inputs, task):
+    """
+    Transcribe or translate an audio input using the Whisper Large V3 model.
+    This function runs automatic speech recognition on an uploaded or recorded
+    audio file using a Hugging Face Transformers pipeline. It supports both
+    transcription (same-language) and translation (to English).
+
+    Args:
+        inputs (str or dict): The audio input, typically a filepath to an audio file
+            provided by Gradio, or a dictionary containing audio array and sampling rate.
+        task (str): The task to perform, either "transcribe" or "translate".
+
+    Returns:
+        str: The transcribed or translated text output from the Whisper model.
+    """
     if inputs is None:
         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):
+    """
+    Generate an HTML iframe embed for a YouTube video.
+    This helper function extracts the video ID from a YouTube URL and
+    returns a simple HTML string for embedding the video in the UI.
+
+    Args:
+        yt_url (str): The full URL of the YouTube video.
+
+    Returns:
+        str: An HTML string containing an iframe embedding the YouTube video.
+    """
     video_id = yt_url.split("?v=")[-1]
     HTML_str = (
         f'<center> <iframe width="500" height="320" src="https://www.youtube.com/embed/{video_id}"> </iframe>'
@@ -41,7 +66,23 @@ def _return_yt_html_embed(yt_url):
     )
     return HTML_str
 
+
 def download_yt_audio(yt_url, filename):
+    """
+    Download a YouTube video's audio to a local file.
+    This function validates the video duration against a maximum length,
+    then downloads the video (with audio) to a specified local filepath
+    using yt-dlp.
+
+    Args:
+        yt_url (str): The URL of the YouTube video to download.
+        filename (str): The destination filepath where the video/audio
+            will be saved.
+
+    Raises:
+        gr.Error: If the video cannot be downloaded or exceeds the
+            maximum allowed duration.
+    """
     info_loader = youtube_dl.YoutubeDL()
     
     try:
@@ -72,8 +113,26 @@ def download_yt_audio(yt_url, filename):
         except youtube_dl.utils.ExtractorError as err:
             raise gr.Error(str(err))
 
+
 @spaces.GPU
 def yt_transcribe(yt_url, task, max_filesize=75.0):
+    """
+    Transcribe or translate audio from a YouTube video.
+    This function downloads a YouTube video, extracts its audio,
+    converts it to the appropriate format, and runs Whisper-based
+    automatic speech recognition on the content.
+
+    Args:
+        yt_url (str): The URL of the YouTube video to transcribe.
+        task (str): The task to perform, either "transcribe" or "translate".
+        max_filesize (float, optional): Maximum allowed file size in MB.
+            Currently unused but kept for interface compatibility.
+
+    Returns:
+        tuple:
+            - html_embed_str (str): An HTML iframe embedding the YouTube video.
+            - text (str): The transcribed or translated text output.
+    """
     html_embed_str = _return_yt_html_embed(yt_url)
 
     with tempfile.TemporaryDirectory() as tmpdirname:
@@ -105,7 +164,6 @@ mf_transcribe = gr.Interface(
         f" checkpoint [{MODEL_NAME}](https://huggingface.co/{MODEL_NAME}) and 🤗 Transformers to transcribe audio files"
         " of arbitrary length."
     ),
-    allow_flagging="never",
 )
 
 file_transcribe = gr.Interface(
@@ -121,7 +179,6 @@ file_transcribe = gr.Interface(
         f" checkpoint [{MODEL_NAME}](https://huggingface.co/{MODEL_NAME}) and 🤗 Transformers to transcribe audio files"
         " of arbitrary length."
     ),
-    allow_flagging="never",
 )
 
 yt_transcribe = gr.Interface(
@@ -137,11 +194,9 @@ yt_transcribe = gr.Interface(
         f" [{MODEL_NAME}](https://huggingface.co/{MODEL_NAME}) and 🤗 Transformers to transcribe video files of"
         " arbitrary length."
     ),
-    allow_flagging="never",
 )
 
 with demo:
     gr.TabbedInterface([mf_transcribe, file_transcribe, yt_transcribe], ["Microphone", "Audio file", "YouTube"])
 
 demo.launch(mcp_server=True)
-