Refactor GPU diffusion model classes for clarity and consistency

- Renamed `DiffusionModelRunner` to `GPUDiffusionModelRunner` for better clarity in class purpose.
- Updated `DiffusionGPUWorker` to utilize the renamed `GPUDiffusionModelRunner`.
- Simplified output handling in `GPUDiffusionModelRunner` by removing unnecessary checks, enhancing code readability.

vllm_omni/worker/gpu_diffusion_worker.py

@@ -9,10 +9,10 @@ from vllm.v1.worker.gpu_worker import init_worker_distributed_environment
 from vllm.model_executor import set_random_seed
 from vllm.v1.utils import report_usage_stats
 
-from vllm_omni.worker.diffusion_model_runner import DiffusionModelRunner
+from vllm_omni.worker.gpu_diffusion_model_runner import GPUDiffusionModelRunner
 
 
-class DiffusionGPUWorker(GPUWorker):
+class GPUDiffusionWorker(GPUWorker):
     def init_device(self):
         if self.device_config.device.type == "cuda":
             # torch.distributed.all_reduce does not free the input tensor until
@@ -59,7 +59,7 @@ class DiffusionGPUWorker(GPUWorker):
         set_random_seed(self.model_config.seed)
 
         # Construct the model runner
-        self.model_runner: DiffusionModelRunner = DiffusionModelRunner(
+        self.model_runner: GPUDiffusionModelRunner = GPUDiffusionModelRunner(
             self.vllm_config, self.device)
 
         if self.rank == 0: