Update handler.py

handler.py

@@ -243,7 +243,7 @@ class EndpointHandler:
         
         # Generate response
         with torch.no_grad():
-            text_streamer = TextStreamer(self.tokenizer)
+            text_streamer = TextStreamer(self.tokenizer,skip_prompt=False)
             output = self.model.generate(
                 **inputs, 
                 streamer=text_streamer,