Update

server/triton.py

@@ -25,10 +25,12 @@ class TritonService:
         )
 
     def generate(self, data: CompletionsRequest) -> List[Choice]:
+        # FIXME(meng): Make following vars configurable
         n = 1
         np_type = np.uint32
         max_tokens = 128
         model_name = "fastertransformer"
+        stop_words = ["\n\n"]
 
         prompt = data.prompt
         input_start_ids = np.expand_dims(self.tokenizer.encode(prompt), 0)
@@ -39,7 +41,6 @@ class TritonService:
         prompt_tokens: int = input_len[0][0]
         output_len = np.ones_like(input_len).astype(np_type) * max_tokens
 
-        stop_words = ["\n\n"]
         stop_word_list = np.repeat(
             to_word_list_format([stop_words], self.tokenizer),
             input_start_ids.shape[0],