[Core] Enable sharded state loader for V1 engine and enhance test coverage (#25308)

Signed-off-by: pengdrumli <pengdrumli@tencent.com>
2025-09-20 21:15:22 +08:00
parent 3c713a9711
commit d88918e4c2
2 changed files with 12 additions and 14 deletions
--- a/vllm/engine/arg_utils.py
+++ b/vllm/engine/arg_utils.py
@@ -1486,12 +1486,6 @@ class EngineArgs:
        #############################################################
        # Unsupported Feature Flags on V1.

-        if self.load_format == "sharded_state":
-            _raise_or_fallback(
-                feature_name=f"--load_format {self.load_format}",
-                recommend_to_remove=False)
-            return False
-
        if (self.logits_processor_pattern
                != EngineArgs.logits_processor_pattern):
            _raise_or_fallback(feature_name="--logits-processor-pattern",