[Feat][RL][1/2] Native Weight Syncing API: NCCL (#31943)

Signed-off-by: ahao-anyscale <ahao@anyscale.com>
Signed-off-by: Aaron Hao <ahao@anyscale.com>
Co-authored-by: SumanthRH <sumanthrh99@gmail.com>

tests/entrypoints/openai/test_openai_schema.py

@@ -139,6 +139,14 @@ def test_openapi_stateless(case: schemathesis.Case):
         # Skip responses API as it is meant to be stateful.
         return
 
+    # Skip weight transfer endpoints as they require special setup
+    # (weight_transfer_config) and are meant to be stateful.
+    if case.operation.path in (
+        "/init_weight_transfer_engine",
+        "/update_weights",
+    ):
+        return
+
     timeout = {
         # requires a longer timeout
         ("POST", "/v1/chat/completions"): LONG_TIMEOUT_SECONDS,