More Multiweight support cleanups (#4948)

* Updated the link for densenet recipe.

* Set default value of `num_classes` and `num_keypoints` to `None`

* Provide helper methods for parameter checks to reduce duplicate code.

* Throw errors on silent config overwrites from weight meta-data and legacy builders.

* Changing order of arguments + fixing mypy.

* Make the builders fully BC.

* Add "default" weights support that returns always the best weights.

test/test_prototype_models.py

@@ -31,14 +31,19 @@ def get_models_with_module_names(module):
 
 
 @pytest.mark.parametrize(
-    "model_fn, weight",
+    "model_fn, name, weight",
     [
-        (models.resnet50, models.ResNet50Weights.ImageNet1K_RefV2),
-        (models.quantization.resnet50, models.quantization.QuantizedResNet50Weights.ImageNet1K_FBGEMM_RefV1),
+        (models.resnet50, "ImageNet1K_RefV1", models.ResNet50Weights.ImageNet1K_RefV1),
+        (models.resnet50, "default", models.ResNet50Weights.ImageNet1K_RefV2),
+        (
+            models.quantization.resnet50,
+            "ImageNet1K_FBGEMM_RefV1",
+            models.quantization.QuantizedResNet50Weights.ImageNet1K_FBGEMM_RefV1,
+        ),
     ],
 )
-def test_get_weight(model_fn, weight):
-    assert models._api.get_weight(model_fn, weight.name) == weight
+def test_get_weight(model_fn, name, weight):
+    assert models._api.get_weight(model_fn, name) == weight
 
 
 @pytest.mark.parametrize("model_fn", TM.get_models_from_module(models))

torchvision/prototype/models/_api.py

@@ -30,6 +30,7 @@ class WeightEntry:
     url: str
     transforms: Callable
     meta: Dict[str, Any]
+    default: bool
 
 
 class Weights(Enum):
@@ -59,7 +60,7 @@ def verify(cls, obj: Any) -> Any:
     @classmethod
     def from_str(cls, value: str) -> "Weights":
         for v in cls:
-            if v._name_ == value:
+            if v._name_ == value or (value == "default" and v.default):
                 return v
         raise ValueError(f"Invalid value {value} for enum {cls.__name__}.")
 

torchvision/prototype/models/_utils.py

@@ -0,0 +1,44 @@
+import warnings
+from typing import Any, Dict, Optional, TypeVar
+
+from ._api import Weights
+
+
+W = TypeVar("W", bound=Weights)
+V = TypeVar("V")
+
+
+def _deprecated_param(
+    kwargs: Dict[str, Any], deprecated_param: str, new_param: str, default_value: Optional[W]
+) -> Optional[W]:
+    warnings.warn(f"The parameter '{deprecated_param}' is deprecated, please use '{new_param}' instead.")
+    if kwargs.pop(deprecated_param):
+        if default_value is not None:
+            return default_value
+        else:
+            raise ValueError("No checkpoint is available for model.")
+    else:
+        return None
+
+
+def _deprecated_positional(kwargs: Dict[str, Any], deprecated_param: str, new_param: str, default_value: V) -> None:
+    warnings.warn(
+        f"The positional parameter '{deprecated_param}' is deprecated, please use keyword parameter '{new_param}'"
+        + " instead."
+    )
+    kwargs[deprecated_param] = default_value
+
+
+def _ovewrite_named_param(kwargs: Dict[str, Any], param: str, new_value: V) -> None:
+    if param in kwargs:
+        if kwargs[param] != new_value:
+            raise ValueError(f"The parameter '{param}' expected value {new_value} but got {kwargs[param]} instead.")
+    else:
+        kwargs[param] = new_value
+
+
+def _ovewrite_value_param(param: Optional[V], new_value: V) -> V:
+    if param is not None:
+        if param != new_value:
+            raise ValueError(f"The parameter '{param}' expected value {new_value} but got {param} instead.")
+    return new_value

torchvision/prototype/models/alexnet.py

@@ -1,4 +1,3 @@
-import warnings
 from functools import partial
 from typing import Any, Optional
 
@@ -8,6 +7,7 @@
 from ...models.alexnet import AlexNet
 from ._api import Weights, WeightEntry
 from ._meta import _IMAGENET_CATEGORIES
+from ._utils import _deprecated_param, _deprecated_positional, _ovewrite_named_param
 
 
 __all__ = ["AlexNet", "AlexNetWeights", "alexnet"]
@@ -25,16 +25,19 @@ class AlexNetWeights(Weights):
             "acc@1": 56.522,
             "acc@5": 79.066,
         },
+        default=True,
     )
 
 
 def alexnet(weights: Optional[AlexNetWeights] = None, progress: bool = True, **kwargs: Any) -> AlexNet:
+    if type(weights) == bool and weights:
+        _deprecated_positional(kwargs, "pretrained", "weights", True)
     if "pretrained" in kwargs:
-        warnings.warn("The parameter pretrained is deprecated, please use weights instead.")
-        weights = AlexNetWeights.ImageNet1K_RefV1 if kwargs.pop("pretrained") else None
+        weights = _deprecated_param(kwargs, "pretrained", "weights", AlexNetWeights.ImageNet1K_RefV1)
     weights = AlexNetWeights.verify(weights)
+
     if weights is not None:
-        kwargs["num_classes"] = len(weights.meta["categories"])
+        _ovewrite_named_param(kwargs, "num_classes", len(weights.meta["categories"]))
 
     model = AlexNet(**kwargs)
 

torchvision/prototype/models/densenet.py

@@ -1,5 +1,4 @@
 import re
-import warnings
 from functools import partial
 from typing import Any, Optional, Tuple
 
@@ -10,6 +9,7 @@
 from ...models.densenet import DenseNet
 from ._api import Weights, WeightEntry
 from ._meta import _IMAGENET_CATEGORIES
+from ._utils import _deprecated_param, _deprecated_positional, _ovewrite_named_param
 
 
 __all__ = [
@@ -53,7 +53,7 @@ def _densenet(
     **kwargs: Any,
 ) -> DenseNet:
     if weights is not None:
-        kwargs["num_classes"] = len(weights.meta["categories"])
+        _ovewrite_named_param(kwargs, "num_classes", len(weights.meta["categories"]))
 
     model = DenseNet(growth_rate, block_config, num_init_features, **kwargs)
 
@@ -67,7 +67,7 @@ def _densenet(
     "size": (224, 224),
     "categories": _IMAGENET_CATEGORIES,
     "interpolation": InterpolationMode.BILINEAR,
-    "recipe": None,  # TODO: add here a URL to documentation stating that the weights were ported from LuaTorch
+    "recipe": "https://github.com/pytorch/vision/pull/116",
 }
 
 
@@ -80,6 +80,7 @@ class DenseNet121Weights(Weights):
             "acc@1": 74.434,
             "acc@5": 91.972,
         },
+        default=True,
     )
 
 
@@ -92,6 +93,7 @@ class DenseNet161Weights(Weights):
             "acc@1": 77.138,
             "acc@5": 93.560,
         },
+        default=True,
     )
 
 
@@ -104,6 +106,7 @@ class DenseNet169Weights(Weights):
             "acc@1": 75.600,
             "acc@5": 92.806,
         },
+        default=True,
     )
 
 
@@ -116,40 +119,45 @@ class DenseNet201Weights(Weights):
             "acc@1": 76.896,
             "acc@5": 93.370,
         },
+        default=True,
     )
 
 
 def densenet121(weights: Optional[DenseNet121Weights] = None, progress: bool = True, **kwargs: Any) -> DenseNet:
+    if type(weights) == bool and weights:
+        _deprecated_positional(kwargs, "pretrained", "weights", True)
     if "pretrained" in kwargs:
-        warnings.warn("The parameter pretrained is deprecated, please use weights instead.")
-        weights = DenseNet121Weights.ImageNet1K_Community if kwargs.pop("pretrained") else None
+        weights = _deprecated_param(kwargs, "pretrained", "weights", DenseNet121Weights.ImageNet1K_Community)
     weights = DenseNet121Weights.verify(weights)
 
     return _densenet(32, (6, 12, 24, 16), 64, weights, progress, **kwargs)
 
 
 def densenet161(weights: Optional[DenseNet161Weights] = None, progress: bool = True, **kwargs: Any) -> DenseNet:
+    if type(weights) == bool and weights:
+        _deprecated_positional(kwargs, "pretrained", "weights", True)
     if "pretrained" in kwargs:
-        warnings.warn("The parameter pretrained is deprecated, please use weights instead.")
-        weights = DenseNet161Weights.ImageNet1K_Community if kwargs.pop("pretrained") else None
+        weights = _deprecated_param(kwargs, "pretrained", "weights", DenseNet161Weights.ImageNet1K_Community)
     weights = DenseNet161Weights.verify(weights)
 
     return _densenet(48, (6, 12, 36, 24), 96, weights, progress, **kwargs)
 
 
 def densenet169(weights: Optional[DenseNet169Weights] = None, progress: bool = True, **kwargs: Any) -> DenseNet:
+    if type(weights) == bool and weights:
+        _deprecated_positional(kwargs, "pretrained", "weights", True)
     if "pretrained" in kwargs:
-        warnings.warn("The parameter pretrained is deprecated, please use weights instead.")
-        weights = DenseNet169Weights.ImageNet1K_Community if kwargs.pop("pretrained") else None
+        weights = _deprecated_param(kwargs, "pretrained", "weights", DenseNet169Weights.ImageNet1K_Community)
     weights = DenseNet169Weights.verify(weights)
 
     return _densenet(32, (6, 12, 32, 32), 64, weights, progress, **kwargs)
 
 
 def densenet201(weights: Optional[DenseNet201Weights] = None, progress: bool = True, **kwargs: Any) -> DenseNet:
+    if type(weights) == bool and weights:
+        _deprecated_positional(kwargs, "pretrained", "weights", True)
     if "pretrained" in kwargs:
-        warnings.warn("The parameter pretrained is deprecated, please use weights instead.")
-        weights = DenseNet201Weights.ImageNet1K_Community if kwargs.pop("pretrained") else None
+        weights = _deprecated_param(kwargs, "pretrained", "weights", DenseNet201Weights.ImageNet1K_Community)
     weights = DenseNet201Weights.verify(weights)
 
     return _densenet(32, (6, 12, 48, 32), 64, weights, progress, **kwargs)