Fix ModuleDict wrapping

stack-info: PR: #260, branch: xmfan/stack/24

Author: xmfan

autoparallel/api.py

@@ -25,7 +25,6 @@
 from torch.distributed.fsdp import MixedPrecisionPolicy
 from torch.distributed.tensor import DeviceMesh
 from torch.export._trace import _restore_state_dict
-from torch.export._unlift import _assign_attr
 from torch.export.unflatten import _AttrKind
 from torch.fx.experimental.symbolic_shapes import ShapeEnv
 
@@ -52,6 +51,66 @@
 _APPLY_VIEW_MM_VIEW_PATTERN = False
 
 
+def _assign_attr(
+    attr: Any,
+    target_module: torch.nn.Module,
+    fqn: str,
+    attr_kind: _AttrKind,
+    ref_module: Optional[torch.nn.Module] = None,
+):
+    """
+    Custom version of torch.export._unlift._assign_attr that preserves the original
+    module structure (e.g., nn.ModuleDict) from ref_module.
+
+    Args:
+        attr: The attribute to assign (parameter/buffer/module)
+        target_module: The module to assign the attribute to
+        fqn: Fully qualified name of the attribute (e.g., "layers.0.weight")
+        attr_kind: Type of attribute (PARAMETER, BUFFER, etc.)
+        ref_module: Reference module to check for original structure (optional)
+    """
+    *prefix, field = fqn.split(".")
+
+    # Navigate to the parent module, creating submodules as needed
+    curr_mod = target_module
+    for i, attr_name in enumerate(prefix):
+        if not hasattr(curr_mod, attr_name):
+            # Check if we should create a ModuleDict or regular Module
+            if ref_module is not None:
+                # Navigate to the same location in ref_module
+                ref_curr_mod = ref_module
+                for ref_attr_name in prefix[:i]:
+                    if hasattr(ref_curr_mod, ref_attr_name):
+                        ref_curr_mod = getattr(ref_curr_mod, ref_attr_name)
+                    else:
+                        ref_curr_mod = None  # type: ignore[assignment]
+                        break
+
+                # Check if the next level should be a ModuleDict
+                if ref_curr_mod is not None and hasattr(ref_curr_mod, attr_name):
+                    ref_submod = getattr(ref_curr_mod, attr_name)
+                    if isinstance(ref_submod, torch.nn.ModuleDict):
+                        setattr(curr_mod, attr_name, torch.nn.ModuleDict())
+                    else:
+                        setattr(curr_mod, attr_name, torch.nn.Module())
+                else:
+                    setattr(curr_mod, attr_name, torch.nn.Module())
+            else:
+                setattr(curr_mod, attr_name, torch.nn.Module())
+
+        curr_mod = getattr(curr_mod, attr_name)
+
+    # Set the final attribute
+    if attr_kind == _AttrKind.PARAMETER:
+        assert isinstance(attr, torch.nn.Parameter)
+        curr_mod.register_parameter(field, attr)
+    elif attr_kind == _AttrKind.BUFFER:
+        assert isinstance(attr, torch.Tensor)
+        curr_mod.register_buffer(field, attr)
+    else:
+        setattr(curr_mod, field, attr)
+
+
 def _get_decomp_table():
     decomp_table = copy.copy(select_decomp_table())
     # TODO: removing those as they cause missing DTensor propagation rules
@@ -550,11 +609,24 @@ def _register_params_and_init_weights(
         # We construct an unflattened structure on parallel_mod,
         # e.g. _assign_attr(v, parallel_model, k="layers.0.weight") will literally
         # create empty nn.Modules recursively and then stash 'v' so it shows up in the right spot
+        # We pass self.model as reference to preserve the original module structure (e.g., nn.ModuleDict)
         for k, v in sharded_param_dict.items():
-            _assign_attr(v, self.parallel_model, k, attr_kind=_AttrKind.PARAMETER)
+            _assign_attr(
+                v,
+                self.parallel_model,
+                k,
+                attr_kind=_AttrKind.PARAMETER,
+                ref_module=self.model,
+            )
 
         for k, v in sharded_buffer_dict.items():
-            _assign_attr(v, self.parallel_model, k, attr_kind=_AttrKind.BUFFER)
+            _assign_attr(
+                v,
+                self.parallel_model,
+                k,
+                attr_kind=_AttrKind.BUFFER,
+                ref_module=self.model,
+            )
 
         # Right now we require a convention that the user model provides an init_weights method,
         # although we could snoop for other methods too.
@@ -621,9 +693,12 @@ def __init__(
         sharded_param_dict: dict[str, torch.nn.Parameter],
         sharded_buffer_dict: dict[str, torch.Tensor],
         init_weights_model: torch.nn.Module,
+        ref_model: torch.nn.Module,
     ):
         super().__init__()
-        self._register_params_and_buffers(sharded_param_dict, sharded_buffer_dict)
+        self._register_params_and_buffers(
+            sharded_param_dict, sharded_buffer_dict, ref_model
+        )
 
         # Right now we require a convention that the user model provides an init_weights method,
         # although we could snoop for other methods too.
@@ -639,16 +714,21 @@ def init_weights(_self, *args, **kwargs):
             # but with our new DTensor sharded params attached to the user module.
             self.init_weights = MethodType(init_weights, self)
 
-    def _register_params_and_buffers(self, sharded_param_dict, sharded_buffer_dict):
+    def _register_params_and_buffers(
+        self, sharded_param_dict, sharded_buffer_dict, ref_model
+    ):
 
         # We construct an unflattened structure on parallel_mod,
         # e.g. _assign_attr(v, parallel_model, k="layers.0.weight") will literally
         # create empty nn.Modules recursively and then stash 'v' so it shows up in the right spot
+        # We pass ref_model to preserve the original module structure (e.g., nn.ModuleDict)
         for k, v in sharded_param_dict.items():
-            _assign_attr(v, self, k, attr_kind=_AttrKind.PARAMETER)
+            _assign_attr(
+                v, self, k, attr_kind=_AttrKind.PARAMETER, ref_module=ref_model
+            )
 
         for k, v in sharded_buffer_dict.items():
-            _assign_attr(v, self, k, attr_kind=_AttrKind.BUFFER)
+            _assign_attr(v, self, k, attr_kind=_AttrKind.BUFFER, ref_module=ref_model)
 
     def forward(self, *args):
         raise NotImplementedError("This is a placeholder for the pipeline model")
@@ -829,6 +909,7 @@ def apply_placement_pp(
             sharded_param_dict,
             sharded_buffer_dict,
             self.init_weights_model,
+            self.model,
         )
         return {
             "graph_callables": graph_modules,

tests/test_api.py

@@ -358,3 +358,62 @@ def input_fn():
         ]
         # Should only have 2 placeholders: weight and input (no tangents for inference)
         assert len(placeholders) == 2
+
+
+def test_moduledict_preservation(device_mesh_1d):
+    """Test that nn.ModuleDict structure is preserved during _assign_attr."""
+    dim = 128
+
+    class Model(nn.Module):
+        def __init__(self, dim):
+            super().__init__()
+            # Create a ModuleDict to test preservation
+            self.layers = nn.ModuleDict(
+                {
+                    "layer1": nn.Linear(dim, dim),
+                    "layer2": nn.Linear(dim, dim),
+                }
+            )
+
+        def forward(self, x):
+            x = self.layers["layer1"](x)
+            x = self.layers["layer2"](x)
+            return x
+
+    def input_fn():
+        b = 512
+        inputs = (torch.rand(b, dim, device="cuda"),)
+        return inputs
+
+    with torch.device("meta"):
+        model = Model(dim)
+
+    # Verify original model has ModuleDict
+    assert isinstance(model.layers, nn.ModuleDict)
+
+    with AutoParallel(
+        model,
+        input_fn,
+        device_mesh_1d,
+    ) as autop:
+        x_sharding = (Shard(0),)
+        autop.add_input_constraints([x_sharding])
+        sharding_placement = autop.optimize_placement()
+
+        # AutoParallel produces a module with meta-DTensor parameters that need to be initialized
+        parallel_mod = autop.apply_placement(sharding_placement)
+
+    # Verify that the parallel_mod preserves the ModuleDict structure
+    assert isinstance(
+        parallel_mod.layers, nn.ModuleDict
+    ), f"Expected nn.ModuleDict but got {type(parallel_mod.layers)}"
+
+    # Verify that the ModuleDict contains the expected layers
+    assert "layer1" in parallel_mod.layers
+    assert "layer2" in parallel_mod.layers
+    assert isinstance(parallel_mod.layers["layer1"], nn.Module)
+    assert isinstance(parallel_mod.layers["layer2"], nn.Module)
+
+    # Verify parameters are accessible through the ModuleDict structure
+    assert hasattr(parallel_mod.layers["layer1"], "weight")
+    assert hasattr(parallel_mod.layers["layer2"], "weight")