Merge branch 'main' into adabeyta/eng_args_profiling

vllm/sequence.py

@@ -6,7 +6,8 @@
 from collections import defaultdict
 from dataclasses import dataclass, field
 from functools import cached_property, reduce
-from typing import TYPE_CHECKING, Any, Callable, Dict, List, Mapping, Optional
+from typing import (TYPE_CHECKING, Any, Callable, DefaultDict, Dict, List,
+                    Mapping, Optional)
 from typing import Sequence as GenericSequence
 from typing import Set, Tuple, Union, cast
 
@@ -256,7 +257,7 @@ def output_token_ids(self) -> Tuple[int, ...]:
         return tuple(self._output_token_ids)
 
     @output_token_ids.setter
-    def output_token_ids(self, new_output_token_ids: List[int]) -> None:
+    def output_token_ids(self, new_output_token_ids: Tuple[int, ...]) -> None:
         self._output_token_ids = array(VLLM_TOKEN_ID_ARRAY_TYPE,
                                        new_output_token_ids)
         self._update_cached_all_tokens()
@@ -1164,7 +1165,7 @@ def get_all_seq_ids_and_request_ids(
     sequence ids.
     """
     seq_ids: List[int] = []
-    request_id_seq_ids_mapping: Dict[str, Set[int]] = defaultdict(set)
+    request_id_seq_ids_mapping: DefaultDict[str, Set[int]] = defaultdict(set)
     for sg in seq_group_metadata_list:
         for seq_id in sg.seq_data:
             seq_ids.append(seq_id)