add chunk_length_s for longer audio

app.py

@@ -69,6 +69,8 @@ def whisper_api(temp_audio_path):
         model=model,
         tokenizer=processor.tokenizer,
         feature_extractor=processor.feature_extractor,
+        chunk_length_s=30,
+        batch_size=32,  # batch size for inference - set based on your device
         torch_dtype=torch_dtype,
         device=device,
     )