Fix edge case for truncated (#1747)

2024-10-23 00:00:25 -04:00
parent 1f26e8b8e4
commit 17536e7e3d
1 changed files with 1 additions and 1 deletions
@@ -416,7 +416,7 @@ class Scheduler:
                )

        # Truncate prompts that are too long
-        if len(req.origin_input_ids) >= self.max_req_input_len:
+        if len(req.origin_input_ids) > self.max_req_input_len:
            logger.warning(
                "Request length is longer than the KV cache pool size or "
                "the max context length. Truncated!!!"