[Export Refactor][Image Classification] create_dummy_input function (…

…#1880)

* initial commit

* looking good, time to cleanup

* Delete src/sparseml/export/helpers.py

* Delete tests/sparseml/export/test_helpers.py

* ready for review

* improve design

* tests pass

* reuse _validate_dataset_num_classes

* initial commit

* Update src/sparseml/pytorch/image_classification/integration_helper_functions.py

* Update src/sparseml/pytorch/image_classification/integration_helper_functions.py

* ready for review

* Update src/sparseml/export/export.py

* Update src/sparseml/integration_helper_functions.py

src/sparseml/export/export.py

@@ -98,9 +98,12 @@ def export(
         IntegrationHelperFunctions.load_from_registry(integration)
     )
 
-    model = helper_functions.create_model(source_path, device)
+    # for now, this code is not runnable, serves as a blueprint
+    model, auxiliary_items = helper_functions.create_model(
+        source_path, **kwargs  # noqa: F821
+    ) 
     sample_data = (
-        helper_functions.create_dummy_input(model, batch_size)
+        helper_functions.create_dummy_input(**auxiliary_items)
         if sample_data is None
         else sample_data
     )

src/sparseml/integration_helper_functions.py

@@ -14,7 +14,7 @@
 
 from enum import Enum
 from pathlib import Path
-from typing import Callable, Optional, Union
+from typing import Any, Callable, Dict, Optional, Union
 
 from pydantic import BaseModel, Field
 
@@ -39,14 +39,27 @@ class IntegrationHelperFunctions(RegistryMixin, BaseModel):
     integration.
     """
 
-    create_model: Optional[Callable] = Field(
-        description="A function that creates a (sparse) "
-        "PyTorch model from a source path and additional "
-        "arguments"
+    create_model: Optional[
+        Callable[
+            [Union[str, Path], Optional[Dict[str, Any]]][
+                "torch.nn.Module", Optional[Dict[str, Any]]  # noqa F821
+            ]
+        ]
+    ] = Field(
+        description="A function that takes: "
+        "- a source path to a PyTorch model "
+        "- (optionally) a dictionary of additional arguments"
+        "and returns: "
+        "- a (sparse) PyTorch model "
+        "- (optionally) a dictionary of additional arguments"
     )
-    create_dummy_input: Optional[Callable] = Field(
-        description="A function that creates a dummy input "
-        "given a (sparse) PyTorch model."
+    create_dummy_input: Optional[
+        Callable[Any]["torch.Tensor"]  # noqa F821
+    ] = Field(  # noqa: F82
+        description="A function that takes: "
+        "- a dictionary of arguments"
+        "and returns: "
+        "- a dummy input for the model (a torch.Tensor) "
     )
     export_model: Optional[Callable] = Field(
         description="A function that exports a (sparse) PyTorch "

src/sparseml/pytorch/image_classification/integration_helper_functions.py

@@ -13,7 +13,7 @@
 # limitations under the License.
 
 from pathlib import Path
-from typing import Any, Union
+from typing import Any, Callable, Dict, Optional, Tuple, Union
 
 import torch
 from pydantic import Field
@@ -27,17 +27,46 @@
 )
 
 
-def create_model(source_path: Union[Path, str], **kwargs) -> torch.nn.Module:
+def create_model(
+    source_path: Union[Path, str], **kwargs
+) -> Tuple[torch.nn.Module, Dict[str, Any]]:
     """
     A contract to create a model from a source path
 
     :param source_path: The path to the model
-    :return: The torch model
+    :param kwargs: Additional kwargs to pass to the model creation function
+    :return: A tuple of the
+        - torch model
+        - additional dictionary of useful objects created during model creation
     """
-    model, *_ = create_image_classification_model(checkpoint_path=source_path, **kwargs)
-    return model
+    model, *_, validation_loader = create_image_classification_model(
+        checkpoint_path=source_path, **kwargs
+    )
+    return model, dict(validation_loader=validation_loader)
+
+
+def create_dummy_input(
+    validation_loader: Optional[torch.utils.data.DataLoader] = None,
+    image_size: Optional[int] = 224,
+) -> torch.Tensor:
+    """
+    A contract to create a dummy input for a model
+
+    :param validation_loader: The validation loader to get a batch from.
+        If None, a fake batch will be created
+    :param image_size: The image size to use for the dummy input. Defaults to 224
+    :return: The dummy input as a torch tensor
+    """
+
+    if not validation_loader:
+        # create fake data for export
+        validation_loader = [[torch.randn(1, 3, image_size, image_size)]]
+    return next(iter(validation_loader))[0]
 
 
 @IntegrationHelperFunctions.register(name=Integrations.image_classification.value)
 class ImageClassification(IntegrationHelperFunctions):
-    create_model: Any = Field(default=create_model)
+    create_model: Callable[..., Tuple[torch.nn.Module, Dict[str, Any]]] = Field(
+        default=create_model
+    )
+    create_dummy_input: Callable[..., torch.Tensor] = Field(default=create_dummy_input)

src/sparseml/pytorch/image_classification/utils/helpers.py

@@ -446,7 +446,7 @@ def create_model(
         if one_shot is not None:
             ScheduledModifierManager.from_yaml(file_path=one_shot).apply(module=model)
 
-        return model, arch_key, checkpoint_path
+        return model, arch_key, checkpoint_path, val_dataset
 
 
 def infer_num_classes(

tests/sparseml/pytorch/image_classification/test_integration_helper_functions.py

@@ -25,3 +25,4 @@ def test_integration_helper_functions():
         Integrations.image_classification.value
     )
     assert image_classification.create_model
+    assert image_classification.create_dummy_input