minor

Author: Your Name

modelopt/torch/opt/conversion.py

@@ -493,10 +493,6 @@ def save(model: nn.Module, f: str | os.PathLike | BinaryIO, **kwargs) -> None:
         model: Any model.
         f: Target file location.
         **kwargs: additional args for ``torch.save()``.
-
-    .. note::
-
-        If model is a wrapper such as DistributedDataParallel, it will be unwrapped for saving.
     """
     # unwrap model
     model = unwrap_model(model, warn=True)
@@ -545,11 +541,6 @@ def restore_from_modelopt_state(model: ModelLike, modelopt_state: dict[str, Any]
     Returns:
         A modified model architecture based on the restored modifications with the unmodified
         weights as stored in the provided ``model`` argument.
-
-    .. note::
-
-        Note that wrappers such as DistributedDataParallel are `not` supported during the restore
-        process. Please wrap the model after the restore process.
     """
     # initialize ModelLikeModule if needed.
     model = model if isinstance(model, nn.Module) else ModelLikeModule(model)
@@ -588,11 +579,6 @@ def restore(model: ModelLike, f: str | os.PathLike | BinaryIO, **kwargs) -> nn.M
 
     Returns:
         The model with original weights and stored architecture.
-
-    .. note::
-
-        Note that wrappers such as DistributedDataParallel are `not` supported during the restore
-        process. Please wrap the model after the restore process.
     """
     # initialize ModelLikeModule if needed.
     model = model if isinstance(model, nn.Module) else ModelLikeModule(model)

modelopt/torch/quantization/calib/histogram.py

@@ -157,8 +157,7 @@ def compute_amax(
         """
         if dist.is_initialized():
             warnings.warn(
-                "This method does not perform any synchronization across DistributedDataParallel"
-                " (DDP) https://pytorch.org/docs/stable/notes/ddp.html modules. The recommended"
+                "This method does not perform any synchronization across distributed processes. The recommended"
                 " method is to use the same calibration dataset across all distributed data"
                 " parallel groups so that `amax` is the same for all DDP modules."
             )