change batch_size & max_new_tokens whisper

app.py

@@ -69,10 +69,10 @@ pipe = pipeline(
     tokenizer=processor.tokenizer,
     feature_extractor=processor.feature_extractor,
     chunk_length_s=30,
-    batch_size=64,  # batch size for inference - set based on your device
+    batch_size=128,  # batch size for inference - set based on your device
     torch_dtype=torch_dtype,
     device=device,
-    max_new_tokens=444,  # 448 max - 4 prompt tokens = 444
+    max_new_tokens=128,  # Limit text generation
     return_timestamps=False,  # Save memory
 )