diff --git a/fastdeploy/model_executor/layers/linear.py b/fastdeploy/model_executor/layers/linear.py
index 91bd5841e..e126aed2b 100644
--- a/fastdeploy/model_executor/layers/linear.py
+++ b/fastdeploy/model_executor/layers/linear.py
@@ -367,10 +367,11 @@ class MergedReplicatedLinear(ReplicatedLinear):
             # loaded_shard_id == "kv_a"
             param_shard_offset = self.output_sizes[0]
             param_shard_size = self.output_sizes[1]
-
+        param_output_dim = True
         if hasattr(param, "tensor_track"):
+            param_output_dim = param.tensor_track.output_dim
             param.tensor_track.mark(start=param_shard_offset, end=param_shard_offset + param_shard_size)
-        param = slice_fn(param, True, start=param_shard_offset, end=param_shard_offset + param_shard_size)
+        param = slice_fn(param, param_output_dim, start=param_shard_offset, end=param_shard_offset + param_shard_size)
         assert param.shape == loaded_weight.shape, (
             f" Attempted to load weight ({loaded_weight.shape}) " f"into parameter ({param.shape})"
         )