[Misc] DeepEPHighThroughtput - Enable Inductor pass (vllm-project#21311)

Signed-off-by: Varun Sundar Rabindranath <[email protected]>
Co-authored-by: Varun Sundar Rabindranath <[email protected]>
Signed-off-by: qizixi <[email protected]>

Author: 2 people

vllm/platforms/cuda.py

@@ -182,9 +182,6 @@ def check_and_update_config(cls, vllm_config: "VllmConfig") -> None:
             compilation_config.use_cudagraph = False
             if model_config is not None:
                 model_config.enforce_eager = True
-            # TODO (varun): Turning this ON gives incorrect results for the
-            # Deepseek-V2-lite model.
-            vllm_config.compilation_config.use_inductor = False
 
     @classmethod
     def get_current_memory_usage(cls,