test

app.py

@@ -5,12 +5,9 @@ import librosa
 import numpy as np
 import soundfile as sf
 import torch
-from flask import Flask, request, jsonify
+import gradio as gr
 from transformers import pipeline
 
-# Initialize Flask app
-app = Flask(__name__)
-
 # Load models globally to avoid reloading on every request
 device = 0 if torch.cuda.is_available() else -1
 models = {
@@ -27,12 +24,12 @@ def load_and_convert_audio(audio_path):
         sf.write(temp_wav.name, audio_data, sample_rate, format='WAV')
         return temp_wav.name
 
-def process_audio(audio_path):
+def process_audio(audio_file):
     """Process audio file and return transcription and summary"""
     results = {}
     
     try:
-        temp_wav_path = load_and_convert_audio(audio_path)
+        temp_wav_path = load_and_convert_audio(audio_file.name)
 
         # Transcription
         transcription = models['transcriber'](temp_wav_path, return_timestamps=True)
@@ -48,7 +45,7 @@ def process_audio(audio_path):
         results['summary'] = ' '.join(summaries)
 
     except Exception as e:
-        return {'error': str(e)}, 500  # Return error message if something goes wrong
+        return {'error': str(e)}  # Return error message if something goes wrong
 
     finally:
         if os.path.exists(temp_wav_path):
@@ -56,21 +53,18 @@ def process_audio(audio_path):
 
     return results
 
+def gradio_interface(audio):
+    """Gradio interface function"""
+    return process_audio(audio)
 
-@app.route('/process-audio', methods=['POST'])
-def process_audio_endpoint():
-    """API endpoint to process audio file"""
-    if 'file' not in request.files:
-        return jsonify({'error': 'No file part'}), 400
-
-    audio_file = request.files['file']
-    temp_audio_path = os.path.join(tempfile.gettempdir(), audio_file.filename)
-    audio_file.save(temp_audio_path)
-
-    results = process_audio(temp_audio_path)
-    os.remove(temp_audio_path)  # Clean up the temporary audio file
-
-    return jsonify(results)
+# Create Gradio interface
+iface = gr.Interface(
+    fn=gradio_interface,
+    inputs=gr.inputs.Audio(source="upload", type="file", label="Upload Audio File"),
+    outputs=["json"],
+    title="Audio Transcription and Summarization",
+    description="Upload an audio file to get its transcription and summary."
+)
 
 if __name__ == "__main__":
-    app.run(host='0.0.0.0', port=5000)
+    iface.launch()

requirements.txt

@@ -1,6 +1,6 @@
-Flask
-transformers
-librosa
+gradio
+torch
 soundfile
+transformers
 numpy
-torch
+flask