Create app.py

app.py

@@ -0,0 +1,51 @@
+import gradio as gr
+from transformers import pipeline
+import os # Import os module to handle file paths
+
+# Load the transcription model
+# Using a smaller model that might fit in memory, or consider running on CPU if memory is still an issue
+# For demonstration, let's try 'openai/whisper-small'
+print("🎤 Loading transcription pipeline...")
+try:
+    transcriber = pipeline("automatic-speech-recognition", model="openai/whisper-small", device="cuda:0") # Explicitly use GPU if available
+except Exception as e:
+    print(f"Could not load model on GPU: {e}. Trying on CPU.")
+    transcriber = pipeline("automatic-speech-recognition", model="openai/whisper-small", device="cpu") # Fallback to CPU
+
+
+def transcribe_audio(audio_file_path):
+    """
+    Transcribes the audio file using the loaded pipeline.
+    """
+    if audio_file_path is None:
+        return "Please upload an audio file."
+
+    # Ensure the file path is accessible
+    if not os.path.exists(audio_file_path):
+        return f"Error: Audio file not found at {audio_file_path}"
+
+    print(f" transcribe the audio file: {audio_file_path}")
+    # Run transcription on the audio file with chunking
+    # Adjust chunk_length_s and return_timestamps as needed for your audio
+    try:
+        transcription_result = transcriber(audio_file_path, chunk_length_s=30, return_timestamps=True)
+        return transcription_result["text"]
+    except Exception as e:
+        return f"Error during transcription: {e}"
+
+
+# Create the Gradio interface
+print("🚀 Creating Gradio interface for Transcription...")
+iface = gr.Interface(
+    fn=transcribe_audio,
+    inputs=gr.Audio(type="filepath", label="Upload Audio File"), # Use type="filepath" to get the path to the temporary file
+    outputs=gr.Textbox(label="Transcription"),
+    title="Audio Transcription Pipeline",
+    description="Upload an audio file (e.g., MP3, WAV) to get a transcription."
+)
+
+# Launch the interface
+print("✨ Launching Gradio interface...")
+# Set share=True to get a public link for sharing (optional)
+iface.launch(debug=True, share=True)
+print("\n✅ Gradio interface launched.")