Fix a bug in runtime backend

python/sglang/lang/interpreter.py

@@ -48,7 +48,10 @@ def run_internal(state, program, func_args, func_kwargs, sync):
 def run_program(
     program, backend, func_args, func_kwargs, default_sampling_para, stream, sync=False
 ):
+    if hasattr(backend, "endpoint"):
+        backend = backend.endpoint
     assert backend is not None, "Please specify a backend"
+
     func_kwargs.update(program.bind_arguments)
     stream_executor = StreamExecutor(
         backend, func_kwargs, default_sampling_para, chat_template=None, stream=stream
@@ -74,6 +77,9 @@ def run_program_batch(
     num_threads,
     progress_bar,
 ):
+    if hasattr(backend, "endpoint"):
+        backend = backend.endpoint
+
     # Extract prefix by tracing and cache it
     if len(batch_arguments) > 1:
         pin_program(program, backend)
@@ -157,9 +163,6 @@ class StreamExecutor:
         self.default_sampling_para = default_sampling_para
         self.stream = stream
 
-        if hasattr(backend, "endpoint"):
-            self.backend = backend.endpoint
-
         self.variables = {}  # Dict[name: str -> value: str]
         self.variable_event = {}  # Dict[name: str -> event: threading.Event]
         self.meta_info = {}  # Dict[name: str -> info: str]