Update app.py

app.py

@@ -12,6 +12,9 @@ import langdetect
 import uuid
 from dotenv import load_dotenv
 import whisper
+from pathlib import Path
+import numpy as np
+from scipy.io import wavfile
 
 load_dotenv()
 
@@ -44,13 +47,25 @@ def transcribe_audio(file_path):
         print("Video file detected. Extracting audio...")
         try:
             video = mp.VideoFileClip(file_path)
-            temp_audio = generate_unique_filename(".mp3")
+            temp_audio = generate_unique_filename(".wav")
             video.audio.write_audiofile(temp_audio)
 
             print(f"temp_audio : {temp_audio}")
             model = whisper.load_model("base")
             print(f"transcription1")
-            transcription = model.transcribe(temp_audio)
+            p = Path(__file__).resolve().parent
+            final_path = p / temp_audio
+            print(final_path)
+            if os.access(str(final_path), os.R_OK):
+                print("File is readable.")
+            else:
+                print("File is not readable. Check permissions.")
+            
+            sample_rate, audio_data = wavfile.read(str(final_path))
+
+
+            transcription = model.transcribe(audio_data, sample_rate=sample_rate)
+            #transcription = model.transcribe(str(final_path))
 
             print(f"transcription {transcription}")
            
@@ -101,10 +116,10 @@ def generate_summary_stream(transcription):
     Please ensure that the summary captures the main points and key ideas of the transcription:
     {transcription[:300000]}..."""
     
-    response, history = model.chat(tokenizer, prompt, history=[])
-    print(f"Final summary generated: {response[:100]}...")
+    #response, history = model.chat(tokenizer, prompt, history=[])
+    #print(f"Final summary generated: {response[:100]}...")
     print("Summary generation completed.")
-    return response
+    return ""
 
 def process_uploaded_video(video_path):
     print(f"Processing uploaded video: {video_path}")