[Fix] Benchmark `"EngineClient" has no attribute "model_config"` (#17976)

Signed-off-by: Brayden Zhong <b8zhong@uwaterloo.ca>

benchmarks/benchmark_throughput.py

@@ -146,9 +146,10 @@ async def run_vllm_async(
 
     async with build_async_engine_client_from_engine_args(
             engine_args, disable_frontend_multiprocessing) as llm:
+        model_config = await llm.get_model_config()
         assert all(
-            llm.model_config.max_model_len >= (request.prompt_len +
-                                               request.expected_output_len)
+            model_config.max_model_len >= (request.prompt_len +
+                                           request.expected_output_len)
             for request in requests), (
                 "Please ensure that max_model_len is greater than the sum of"
                 " prompt_len and expected_output_len for all requests.")
@@ -599,7 +600,7 @@ if __name__ == "__main__":
         "--lora-path",
         type=str,
         default=None,
-        help="Path to the lora adapters to use. This can be an absolute path, "
+        help="Path to the LoRA adapters to use. This can be an absolute path, "
         "a relative path, or a Hugging Face model identifier.")
     parser.add_argument(
         "--prefix-len",

vllm/benchmarks/throughput.py

@@ -148,9 +148,10 @@ async def run_vllm_async(
 
     async with build_async_engine_client_from_engine_args(
             engine_args, disable_frontend_multiprocessing) as llm:
+        model_config = await llm.get_model_config()
         assert all(
-            llm.model_config.max_model_len >= (request.prompt_len +
-                                               request.expected_output_len)
+            model_config.max_model_len >= (request.prompt_len +
+                                           request.expected_output_len)
             for request in requests), (
                 "Please ensure that max_model_len is greater than the sum of"
                 " prompt_len and expected_output_len for all requests.")