fix(bindings): serve_endpoint no longer takes a lease (#1014)

docs/guides/dynamo_run.md

@@ -438,7 +438,7 @@ async def worker(runtime: DistributedRuntime):
 
     # 3. Attach request handler
     #
-    await endpoint.serve_endpoint(RequestHandler(engine).generate, None)
+    await endpoint.serve_endpoint(RequestHandler(engine).generate)
 
 class RequestHandler:
 

launch/dynamo-run/src/subprocess/sglang_inc.py

@@ -134,7 +134,7 @@ async def init(runtime: DistributedRuntime, config: Config):
 
     # the server will gracefully shutdown (i.e., keep opened TCP streams finishes)
     # after the lease is revoked
-    await endpoint.serve_endpoint(RequestHandler(engine_client).generate, None)
+    await endpoint.serve_endpoint(RequestHandler(engine_client).generate)
 
 
 def cmd_line_args():

launch/dynamo-run/src/subprocess/vllm_inc.py

@@ -154,7 +154,7 @@ async def init(runtime: DistributedRuntime, config: Config):
     # the server will gracefully shutdown (i.e., keep opened TCP streams finishes)
     # after the lease is revoked
     await endpoint.serve_endpoint(
-        RequestHandler(engine_client, default_sampling_params).generate, None
+        RequestHandler(engine_client, default_sampling_params).generate
     )
 
 

lib/bindings/python/examples/hello_world/server_sglang.py

@@ -108,7 +108,7 @@ async def init(runtime: DistributedRuntime, config: Config):
 
     # the server will gracefully shutdown (i.e., keep opened TCP streams finishes)
     # after the lease is revoked
-    await endpoint.serve_endpoint(RequestHandler(engine_client).generate, None)
+    await endpoint.serve_endpoint(RequestHandler(engine_client).generate)
 
 
 def cmd_line_args():

lib/bindings/python/examples/hello_world/server_vllm.py

@@ -125,7 +125,7 @@ async def init(runtime: DistributedRuntime, config: Config):
 
     # the server will gracefully shutdown (i.e., keep opened TCP streams finishes)
     # after the lease is revoked
-    await endpoint.serve_endpoint(RequestHandler(engine_client).generate, None)
+    await endpoint.serve_endpoint(RequestHandler(engine_client).generate)
 
 
 def cmd_line_args():