Update app.py

app.py

@@ -63,22 +63,22 @@ def transcribe_speech(filepath):
     # Check length and process
     if input_features.shape[-1] > 3000:
         print("Splitting audio required")
-        from pydub import AudioSegment
-
-        def split_audio(file_path, chunk_length_ms=30000):  # 30 sec chunks
-            audio = AudioSegment.from_file(file_path)
-            chunks = [audio[i:i+chunk_length_ms] for i in range(0, len(audio), chunk_length_ms)]
-            return chunks
-
-        # Split and transcribe
-        audio_chunks = split_audio(audio_path)
-
-        for i, chunk in enumerate(audio_chunks):
-            print(i)
-            chunk.export(f"chunk_{i}.wav", format="wav")
-            result = pipe(f"chunk_{i}.wav")
-            output += result['text'] + " "
-            print(f"Chunk {i}: {result['text']}")
+        # from pydub import AudioSegment
+
+        # def split_audio(file_path, chunk_length_ms=30000):  # 30 sec chunks
+        #     audio = AudioSegment.from_file(file_path)
+        #     chunks = [audio[i:i+chunk_length_ms] for i in range(0, len(audio), chunk_length_ms)]
+        #     return chunks
+
+        # # Split and transcribe
+        # audio_chunks = split_audio(audio_path)
+
+        # for i, chunk in enumerate(audio_chunks):
+        #     print(i)
+        #     chunk.export(f"chunk_{i}.wav", format="wav")
+        #     result = pipe(f"chunk_{i}.wav")
+        #     output += result['text'] + " "
+        #     print(f"Chunk {i}: {result['text']}")
     else:
         predicted_ids = model.generate(input_features)
         transcription = processor.batch_decode(predicted_ids, skip_special_tokens=True)