From 1a4baaf89272f805db7a4a1bb50f6bfcb6e130ed Mon Sep 17 00:00:00 2001
From: shuyix <shuyix@nvidia.com>
Date: Thu, 24 Jul 2025 02:29:11 -0700
Subject: [PATCH 1/2] enable partial load

---
 tensorrt_llm/_torch/models/modeling_utils.py | 20 +++++++++++++-------
 1 file changed, 13 insertions(+), 7 deletions(-)

diff --git a/tensorrt_llm/_torch/models/modeling_utils.py b/tensorrt_llm/_torch/models/modeling_utils.py
index a8ce31bf2cef..17dc332eb2f9 100755
--- a/tensorrt_llm/_torch/models/modeling_utils.py
+++ b/tensorrt_llm/_torch/models/modeling_utils.py
@@ -724,6 +724,9 @@ def load_single_module(name, module):
                 for new_name in params_map[names[-1]]:
                     fw = filter_weights('.'.join(names[:-1] + [new_name]),
                                         weights)
+                    # tmp fixes to enable partial updates in old path
+                    if not fw:
+                        continue
                     if new_name in ['k_proj', 'v_proj']:
                         num_kv_heads_list = [num_kv_heads
                                              ] * len(fw) if isinstance(
@@ -740,15 +743,18 @@ def load_single_module(name, module):
                         }
 
                     module_weights.append(fw)
-                module.load_weights(weights=module_weights)
+                if module_weights:
+                    module.load_weights(weights=module_weights)
+
             else:
                 module_weights = filter_weights(name, weights)
-                if hasattr(module, 'load_weights'):
-                    module.load_weights(weights=[module_weights])
-                else:
-                    for n, p in module._parameters.items():
-                        if p is not None:
-                            p.data.copy_(module_weights[n][:])
+                if module_weights:
+                    if hasattr(module, 'load_weights'):
+                        module.load_weights(weights=[module_weights])
+                    else:
+                        for n, p in module._parameters.items():
+                            if p is not None:
+                                p.data.copy_(module_weights[n][:])
 
     if os.environ.get("TRT_LLM_DISABLE_LOAD_WEIGHTS_IN_PARALLEL",
                       False) in ["True", "true", "1", "yes", "y"]:

From 98359938c091be23561cb6c2cb55f1bb5be1aa8c Mon Sep 17 00:00:00 2001
From: shuyix <shuyix@nvidia.com>
Date: Thu, 24 Jul 2025 02:30:04 -0700
Subject: [PATCH 2/2] align interfaces with ray branch

---
 tensorrt_llm/_torch/pyexecutor/py_executor.py | 57 +++++++++++++------
 1 file changed, 39 insertions(+), 18 deletions(-)

diff --git a/tensorrt_llm/_torch/pyexecutor/py_executor.py b/tensorrt_llm/_torch/pyexecutor/py_executor.py
index 30e2a8a16c3d..0d7382c9377d 100644
--- a/tensorrt_llm/_torch/pyexecutor/py_executor.py
+++ b/tensorrt_llm/_torch/pyexecutor/py_executor.py
@@ -1084,6 +1084,44 @@ def _prepare_draft_requests(self):
             logger.error(f"Encountered an error in decode: {error_msg}")
             self._handle_errors(error_msg)
 
+    def update_weights(self, weights):
+        # Load weights into the model
+        self.model_engine.model.load_weights(weights)
+        torch.cuda.synchronize()
+
+        # TODO: reset prefix cache
+
+    def update_weight_from_ipc_handles(self, handles):
+        """
+        Update model weights from IPC handles.
+        
+        Args:
+            ipc_handles (dict): Dictionary mapping device UUIDs to parameter IPC handles.
+                {device_uuid: all_handles}
+        """
+        from tensorrt_llm._torch.utils import get_device_uuid
+        device_uuid = get_device_uuid(self.device_id)
+        
+        if device_uuid not in handles:
+            raise ValueError(f"Device UUID {device_uuid} not found in ipc_handles")
+            
+        try:
+            weights = {}
+            all_handles = handles[device_uuid]
+
+            for param_name, tensor_handle in all_handles:
+                func, args = tensor_handle
+                list_args = list(args)
+                list_args[6] = self.device_id  # Set target device
+                tensor = func(*list_args)
+                weights[param_name] = tensor
+
+            self.update_weights(weights)
+                
+        except Exception as e:
+            logger.error(f"failed to update weights from ipc handles: {e}")
+            return False
+
     def _sleep(self, sleep_request):
         self.is_sleep_request = False
         self._enqueue_responses({sleep_request.id: LlmResponse(request_id=sleep_request.id, result=LlmResult(result=None, py_result=PyResult(0, 0, success=True), is_final=True), client_id=sleep_request.id)})
@@ -1096,24 +1134,7 @@ def _update_weight(self, update_weight_request):
         self.is_update_weight_request = False
 
         try:
-            # Get handles for this device
-            device_uuid = get_device_uuid(self.device_id)
-            handles = update_weight_request.weight_ipc_handles[device_uuid]
-            weights = {}
-
-            # Process each handle to get the tensor
-            for name, handle in handles:
-                func, args = handle
-                list_args = list(args)
-                # Update device ID to match the current device
-                list_args[6] = self.device_id
-                tensor = func(*list_args)
-                weights[name] = tensor
-
-            # Load weights into the model
-            self.model_engine.model.load_weights(weights)
-
-            torch.cuda.synchronize()
+            self.update_weight_from_ipc_handles(update_weight_request.weight_ipc_handles)
             update_weight_response = LlmResponse(request_id=update_weight_request.id, result=LlmResult(result=None, py_result=PyResult(0, 0, success=True), is_final=True),     client_id=update_weight_request.id)
             self._enqueue_responses({update_weight_request.id: update_weight_response})
         except Exception as e: