[core] [2/N] refactor worker_base input preparation for multi-step (#7387)

vllm/worker/worker.py

@@ -264,6 +264,7 @@ class Worker(LocalOrDistributedWorkerBase):
     def prepare_worker_input(
             self, execute_model_req: ExecuteModelRequest) -> WorkerInput:
         virtual_engine = execute_model_req.virtual_engine
+        num_steps = execute_model_req.num_steps
         num_seq_groups = len(execute_model_req.seq_group_metadata_list)
         # `blocks_to_swap_in` and `blocks_to_swap_out` are cpu tensors.
         # they contain parameters to launch cudamemcpyasync.
@@ -286,6 +287,7 @@ class Worker(LocalOrDistributedWorkerBase):
             blocks_to_swap_out=blocks_to_swap_out,
             blocks_to_copy=blocks_to_copy,
             virtual_engine=virtual_engine,
+            num_steps=num_steps,
         )
 
     @torch.inference_mode()

vllm/worker/worker_base.py

@@ -129,6 +129,7 @@ class WorkerInput:
     blocks_to_swap_out: Optional[torch.Tensor] = None
     blocks_to_copy: Optional[torch.Tensor] = None
     virtual_engine: int = 0
+    num_steps: int = 1
 
     @classmethod
     def from_broadcasted_tensor_dict(
@@ -145,6 +146,7 @@ class WorkerInput:
             blocks_to_swap_out=tensor_dict.pop("blocks_to_swap_out"),
             blocks_to_copy=tensor_dict.pop("blocks_to_copy"),
             virtual_engine=tensor_dict["virtual_engine"],
+            num_steps=tensor_dict.pop("num_steps"),
         )
 
     def as_broadcastable_tensor_dict(
@@ -158,6 +160,7 @@ class WorkerInput:
             "blocks_to_swap_out": self.blocks_to_swap_out,
             "blocks_to_copy": self.blocks_to_copy,
             "virtual_engine": self.virtual_engine,
+            "num_steps": self.num_steps,
         }
 
         return tensor_dict
@@ -216,13 +219,50 @@ class LocalOrDistributedWorkerBase(WorkerBase):
         """
         raise NotImplementedError
 
-    def execute_model(
+    def _get_worker_input_from_broadcast(
+            self) -> Optional[Tuple[ModelRunnerInputBase, WorkerInput]]:
+        """ Get the worker input from the broadcasted tensor dict. """
+        assert self.do_metadata_broadcast
+        assert not self.is_driver_worker
+        broadcast_data = broadcast_tensor_dict(src=0)
+        if not broadcast_data:
+            return None
+
+        worker_input = WorkerInput.from_broadcasted_tensor_dict(broadcast_data)
+        model_input = (
+            self.model_runner.make_model_input_from_broadcasted_tensor_dict(
+                broadcast_data))
+
+        return model_input, worker_input
+
+    def _get_driver_input_and_broadcast(
+        self, execute_model_req: ExecuteModelRequest
+    ) -> Tuple[ModelRunnerInputBase, WorkerInput]:
+        """ Get the driver input and broadcast it to other workers.  """
+        assert self.is_driver_worker
+
+        worker_input: WorkerInput = self.prepare_worker_input(
+            execute_model_req=execute_model_req)
+        model_input: ModelRunnerInputBase = (
+            self.model_runner.prepare_model_input(
+                execute_model_req.seq_group_metadata_list,
+                execute_model_req.virtual_engine,
+                execute_model_req.finished_requests_ids))
+
+        if self.do_metadata_broadcast:
+            broadcast_data = worker_input.as_broadcastable_tensor_dict()
+            broadcast_data.update(model_input.as_broadcastable_tensor_dict())
+            broadcast_tensor_dict(broadcast_data, src=0)
+
+        return model_input, worker_input
+
+    def prepare_input(
         self,
         execute_model_req: Optional[ExecuteModelRequest] = None
-    ) -> Optional[List[SamplerOutput]]:
+    ) -> Optional[Tuple[ModelRunnerInputBase, WorkerInput]]:
-        """Executes at least one model step on the given sequences, unless no
+        """
-        sequences are provided."""
+        Prepare the inputs to ModelRunner and workers.
-        start_time = time.perf_counter()
+        """
         if self.is_driver_worker:
             if execute_model_req is None:
                 if self.do_metadata_broadcast:
@@ -233,34 +273,24 @@ class LocalOrDistributedWorkerBase(WorkerBase):
                     # notify all other workers to stop their execution loop.
                     broadcast_tensor_dict({}, src=0)
                 return None
-
+            return self._get_driver_input_and_broadcast(execute_model_req)
-            worker_input: WorkerInput = self.prepare_worker_input(
-                execute_model_req=execute_model_req)
-            model_input: ModelRunnerInputBase = (
-                self.model_runner.prepare_model_input(
-                    execute_model_req.seq_group_metadata_list,
-                    execute_model_req.virtual_engine,
-                    execute_model_req.finished_requests_ids))
-            num_steps = execute_model_req.num_steps
-
-            if self.do_metadata_broadcast:
-                broadcast_data = worker_input.as_broadcastable_tensor_dict()
-                broadcast_data.update(
-                    model_input.as_broadcastable_tensor_dict())
-                broadcast_data["num_steps"] = num_steps
-                broadcast_tensor_dict(broadcast_data, src=0)
         else:
-            assert self.do_metadata_broadcast
+            return self._get_worker_input_from_broadcast()
-            broadcast_data = broadcast_tensor_dict(src=0)
-            if not broadcast_data:
-                return None
 
-            num_steps = broadcast_data.pop("num_steps")
+    def execute_model(
-            worker_input = WorkerInput.from_broadcasted_tensor_dict(
+        self,
-                broadcast_data)
+        execute_model_req: Optional[ExecuteModelRequest] = None
-            model_input = (
+    ) -> Optional[List[SamplerOutput]]:
-                self.model_runner.
+        """Executes at least one model step on the given sequences, unless no
-                make_model_input_from_broadcasted_tensor_dict(broadcast_data))
+        sequences are provided."""
+        start_time = time.perf_counter()
+
+        inputs = self.prepare_input(execute_model_req)
+        if inputs is None:
+            return None
+
+        model_input, worker_input = inputs
+        num_steps = worker_input.num_steps
 
         self.execute_worker(worker_input)