fix stop string

Signed-off-by: Ceng23333 <441651826@qq.com>

python/infinilm/llm/llm.py

@@ -246,14 +246,26 @@ class LLMEngine:
                     req.mark_finished(req.finish_reason)
                     finished_now = True
 
+                    # Remove stop string from generated_text if STOP_STRING finish reason
+                    if req.finish_reason == FinishReason.STOP_STRING:
+                        stop_strings = req.sampling_params.stop or []
+                        for stop_str in stop_strings:
+                            if decoded_text.endswith(stop_str):
+                                # Remove the stop string from the end
+                                decoded_text = decoded_text[:-len(stop_str)]
+                                req.generated_text = decoded_text
+                                break
+
                 holds_back_incomplete_utf8 = (
                     bool(decoded_text) and decoded_text.endswith("\ufffd")
                 )
 
                 # vLLM-style: hold back only if we are not on the final chunk.
-                # Also suppress output when finish reason is LENGTH to avoid replacement issues.
+                # Suppress output when finish reason is LENGTH or STOP_STRING.
+                # Root cause fix: When STOP_STRING is detected, we suppress output for the token
+                # that completes the stop string, preventing additional tokens from being output.
                 if (holds_back_incomplete_utf8 and not finished_now) or (
-                    finished_now and req.finish_reason == FinishReason.LENGTH
+                    finished_now and req.finish_reason in (FinishReason.LENGTH, FinishReason.STOP_STRING)
                 ):
                     token_text = ""
                 else:
@@ -265,6 +277,14 @@ class LLMEngine:
             # For non-streaming, finish checks happen here.
             if req._output_queue is None and self._check_request_finished(req, token_id):
                 req.mark_finished(req.finish_reason)
+                # Remove stop string from generated_text if STOP_STRING finish reason
+                if req.finish_reason == FinishReason.STOP_STRING:
+                    stop_strings = req.sampling_params.stop or []
+                    for stop_str in stop_strings:
+                        if req.generated_text.endswith(stop_str):
+                            # Remove the stop string from the end
+                            req.generated_text = req.generated_text[:-len(stop_str)]
+                            break
 
             # Put output in queue if it exists (for async streaming)
             if req._output_queue is not None: