Simplify guard

Author: amorehead

src/lightning/pytorch/plugins/precision/fsdp.py

@@ -84,7 +84,7 @@ def convert_module(self, module: Module) -> Module:
     @override
     def clip_grad_by_norm(self, module: Optional[Module], optimizer: Optimizer, clip_val: Union[int, float]) -> None:
         # see https://pytorch.org/docs/stable/fsdp.html#torch.distributed.fsdp.FullyShardedDataParallel.clip_grad_norm_
-        if module is None or not hasattr(module, "clip_grad_norm_") or not isinstance(module.clip_grad_norm_, callable):
+        if module is None:
             return
         module.clip_grad_norm_(clip_val)