[Bugfix] Fix config misalignment between offline and online diffusion inference (Wan2.2, Qwen-Image series) (#1979)

Signed-off-by: samithuang <285365963@qq.com>
Signed-off-by: Samit <285365963@qq.com>
Co-authored-by: Hongsheng Liu <liuhongsheng4@huawei.com>

vllm_omni/diffusion/worker/diffusion_model_runner.py

@@ -221,6 +221,7 @@ class DiffusionModelRunner:
                 not getattr(req, "skip_cache_refresh", False)
                 and self.cache_backend is not None
                 and self.cache_backend.is_enabled()
+                and req.sampling_params.num_inference_steps is not None
             ):
                 self.cache_backend.refresh(self.pipeline, req.sampling_params.num_inference_steps)