Update app.py

app.py

@@ -1,53 +1,54 @@
 import torch
 from transformers import pipeline
-from transformers.pipelines.audio_utils import ffmpeg_read
 import gradio as gr
 
-MODEL_NAME = "riteshkr/quantized-whisper-large-v3"
+# Define the model details
+MODEL_NAME = "riteshkr/quantized-whisper-large-v3"  # Update with your actual model ID
 BATCH_SIZE = 8
 
+# Select device based on availability of CUDA (GPU) or fallback to CPU
 device = 0 if torch.cuda.is_available() else "cpu"
 
+# Load the ASR model pipeline
 pipe = pipeline(
     task="automatic-speech-recognition",
     model=MODEL_NAME,
     chunk_length_s=30,
     device=device,
 )
+
+# Utility function to format timestamps
 def format_timestamp(seconds: float, always_include_hours: bool = False, decimal_marker: str = "."):
     if seconds is not None:
         milliseconds = round(seconds * 1000.0)
-
         hours = milliseconds // 3_600_000
         milliseconds -= hours * 3_600_000
-
         minutes = milliseconds // 60_000
         milliseconds -= minutes * 60_000
-
         seconds = milliseconds // 1_000
         milliseconds -= seconds * 1_000
-
         hours_marker = f"{hours:02d}:" if always_include_hours or hours > 0 else ""
         return f"{hours_marker}{minutes:02d}:{seconds:02d}{decimal_marker}{milliseconds:03d}"
     else:
         return seconds
 
+# Transcription function for batch processing
+def transcribe(files, task, return_timestamps):
+    transcriptions = []
+    for file in files:  # Process each file in the batch
+        outputs = pipe(file, batch_size=BATCH_SIZE, generate_kwargs={"task": task}, return_timestamps=return_timestamps)
+        text = outputs["text"]
+        if return_timestamps:
+            timestamps = outputs["chunks"]
+            formatted_chunks = [
+                f"[{format_timestamp(chunk['timestamp'][0])} -> {format_timestamp(chunk['timestamp'][1])}] {chunk['text']}"
+                for chunk in timestamps
+            ]
+            text = "\n".join(formatted_chunks)
+        transcriptions.append(text)
+    return "\n\n".join(transcriptions)  # Return all transcriptions combined
 
-def transcribe(file, task, return_timestamps):
-    outputs = pipe(file, batch_size=BATCH_SIZE, generate_kwargs={"task": task}, return_timestamps=return_timestamps)
-    text = outputs["text"]
-    if return_timestamps:
-        timestamps = outputs["chunks"]
-        timestamps = [
-            f"[{format_timestamp(chunk['timestamp'][0])} -> {format_timestamp(chunk['timestamp'][1])}] {chunk['text']}"
-            for chunk in timestamps
-        ]
-        text = "\n".join(str(feature) for feature in timestamps)
-    return text
-
-
-demo = gr.Blocks()
-
+# Define Gradio interface for microphone input
 mic_transcribe = gr.Interface(
     fn=transcribe,
     inputs=[
@@ -57,44 +58,43 @@ mic_transcribe = gr.Interface(
     ],
     outputs="text",
     layout="horizontal",
-    theme="huggingface",
-    title="Transcribe Audio",
+    title="Whisper Demo: Transcribe Audio",
     description=(
-        "Transcribe long-form microphone or audio inputs with the click of a button! Demo uses the"
-        f" checkpoint [{MODEL_NAME}](https://huggingface.co/{MODEL_NAME}) and 🤗 Transformers to transcribe audio files"
-        " of arbitrary length."
+        f"Transcribe long-form microphone inputs with the {MODEL_NAME} model. Supports transcription and translation."
     ),
     allow_flagging="never",
 )
 
+# Define Gradio interface for file upload
 file_transcribe = gr.Interface(
     fn=transcribe,
     inputs=[
-        gr.Audio(sources="upload", label="Audio file", type="filepath"),
+        gr.Audio(sources="upload", type="filepath", label="Upload Audio File"),
         gr.Radio(["transcribe", "translate"], label="Task"),
         gr.Checkbox(label="Return timestamps"),
     ],
     outputs="text",
     layout="horizontal",
-    theme="huggingface",
-    title="Transcribe Audio",
+    title="Whisper Demo: Transcribe Audio",
     description=(
-        "Transcribe long-form microphone or audio inputs with the click of a button! Demo uses the"
-        f" checkpoint [{MODEL_NAME}](https://huggingface.co/{MODEL_NAME}) and 🤗 Transformers to transcribe audio files"
-        " of arbitrary length."
+        f"Upload audio files to transcribe or translate them using the {MODEL_NAME} model."
     ),
+    allow_flagging="never",
     examples=[
         ["./example.flac", "transcribe", False],
         ["./example.flac", "transcribe", True],
     ],
-    cache_examples=True,
-    allow_flagging="never",
 )
 
+# Create the Gradio tabbed interface for switching between modes
+demo = gr.Blocks()
+
 with demo:
     gr.TabbedInterface(
-        [mic_transcribe, file_transcribe], 
+        [mic_transcribe, file_transcribe],
         ["Transcribe Microphone", "Transcribe Audio File"]
     )
 
-demo.launch(enable_queue=True)
+# Launch the app
+if __name__ == "__main__":
+    demo.launch(debug=True, enable_queue=True, share=True)