diff --git a/xtuner/v1/config/fsdp.py b/xtuner/v1/config/fsdp.py
index c7341799a..99d0d1202 100644
--- a/xtuner/v1/config/fsdp.py
+++ b/xtuner/v1/config/fsdp.py
@@ -22,6 +22,7 @@ class FSDPConfig(BaseModel):
     # TODO: (caoweihan) Convert `torch.dtype` to `Annotated` for compatibility with cyclopts
     param_dtype: Annotated[torch.dtype, Parameter(help="Data type for model parameters")] = torch.bfloat16
     reduce_dtype: Annotated[torch.dtype, Parameter(help="Data type for reduction operations")] = torch.bfloat16
+    lm_head_fp32: Annotated[bool, Parameter(help="Use float32 for language model head")] = False
     torch_compile: Annotated[bool, Parameter(help="Enable model compilation for faster inference")] = False
     compile_targets: Annotated[
         Optional[Tuple[str, ...]],
diff --git a/xtuner/v1/model/dense/dense.py b/xtuner/v1/model/dense/dense.py
index 70839d241..e5a4e26de 100644
--- a/xtuner/v1/model/dense/dense.py
+++ b/xtuner/v1/model/dense/dense.py
@@ -186,6 +186,10 @@ def fully_shard(
         mp_policy = MixedPrecisionPolicy(
             param_dtype=self.fsdp_config.param_dtype, reduce_dtype=fsdp_config.reduce_dtype
         )
+        if self.fsdp_config.lm_head_fp32:
+            lm_head_mp_policy = MixedPrecisionPolicy(param_dtype=torch.float32, reduce_dtype=torch.float32)
+        else:
+            lm_head_mp_policy = mp_policy
         num_recompute_layers = int(self.config.num_hidden_layers * self.fsdp_config.recompute_ratio)
 
         generator = torch.Generator()
@@ -236,7 +240,7 @@ def fully_shard(
         fully_shard(
             self.lm_head,
             mesh=self.fsdp_mesh if self.hsdp_mesh is None else self.hsdp_mesh,
-            mp_policy=mp_policy,
+            mp_policy=lm_head_mp_policy,
             reshard_after_forward=self.fsdp_config.reshard_after_forward,
             offload_policy=CPUOffloadPolicy() if self.fsdp_config.cpu_offload else None,
         )
diff --git a/xtuner/v1/model/moe/moe.py b/xtuner/v1/model/moe/moe.py
index c5703b5fb..ef8a6f43e 100644
--- a/xtuner/v1/model/moe/moe.py
+++ b/xtuner/v1/model/moe/moe.py
@@ -653,6 +653,10 @@ def fully_shard(
         mp_policy = MixedPrecisionPolicy(
             param_dtype=self.fsdp_config.param_dtype, reduce_dtype=fsdp_config.reduce_dtype
         )
+        if self.fsdp_config.lm_head_fp32:
+            lm_head_mp_policy = MixedPrecisionPolicy(param_dtype=torch.float32, reduce_dtype=torch.float32)
+        else:
+            lm_head_mp_policy = mp_policy
         num_recompute_layers = int(self.config.num_hidden_layers * self.fsdp_config.recompute_ratio)
 
         for layer_idx, layer in tqdm(self.layers.items(), desc="[FSDP Sharding]"):
@@ -698,7 +702,7 @@ def fully_shard(
         fully_shard(
             self.lm_head,
             mesh=self.fsdp_mesh if self.hsdp_mesh is None else self.hsdp_mesh,
-            mp_policy=mp_policy,
+            mp_policy=lm_head_mp_policy,
             reshard_after_forward=self.fsdp_config.reshard_after_forward,
             offload_policy=CPUOffloadPolicy() if self.fsdp_config.cpu_offload else None,
         )