[Bugfix][Model] Make Olmo2Model weight loading return loaded weights (vllm-project#18504)

Signed-off-by: Shane A <[email protected]>
Signed-off-by: Yuqi Zhang <[email protected]>

Author: 2015aroras

vllm/model_executor/models/olmo2.py

@@ -314,7 +314,8 @@ def forward(
         hidden_states = self.norm(hidden_states)
         return hidden_states
 
-    def load_weights(self, weights: Iterable[tuple[str, torch.Tensor]]):
+    def load_weights(self, weights: Iterable[tuple[str,
+                                                   torch.Tensor]]) -> set[str]:
         stacked_params_mapping = [
             # (param_name, shard_name, shard_id)
             ("qkv_proj", "q_proj", "q"),
@@ -325,6 +326,7 @@ def load_weights(self, weights: Iterable[tuple[str, torch.Tensor]]):
         ]
 
         params_dict = dict(self.named_parameters(remove_duplicate=False))
+        loaded_params: set[str] = set()
         for name, loaded_weight in weights:
             if is_pp_missing_parameter(name, self):
                 continue
@@ -347,6 +349,8 @@ def load_weights(self, weights: Iterable[tuple[str, torch.Tensor]]):
                 weight_loader = getattr(param, "weight_loader",
                                         default_weight_loader)
                 weight_loader(param, loaded_weight)
+            loaded_params.add(name)
+        return loaded_params
 
 
 class Olmo2ForCausalLM(nn.Module, SupportsPP):