Highlighting the Note in the docs

torchvision/models/quantization/googlenet.py

@@ -143,9 +143,10 @@ def googlenet(
 ) -> QuantizableGoogLeNet:
     """GoogLeNet (Inception v1) model architecture from `Going Deeper with Convolutions <http://arxiv.org/abs/1409.4842>`__.
 
-    Note that ``quantize = True`` returns a quantized model with 8 bit
-    weights. Quantized models only support inference and run on CPUs.
-    GPU inference is not yet supported
+    .. note::
+        Note that ``quantize = True`` returns a quantized model with 8 bit
+        weights. Quantized models only support inference and run on CPUs.
+        GPU inference is not yet supported.
 
     Args:
         weights (:class:`~torchvision.models.quantization.GoogLeNet_QuantizedWeights` or :class:`~torchvision.models.GoogLeNet_Weights`, optional): The

torchvision/models/quantization/inception.py

@@ -214,9 +214,10 @@ def inception_v3(
         **Important**: In contrast to the other models the inception_v3 expects tensors with a size of
         N x 3 x 299 x 299, so ensure your images are sized accordingly.
 
-    Note that quantize = True returns a quantized model with 8 bit
-    weights. Quantized models only support inference and run on CPUs.
-    GPU inference is not yet supported
+    .. note::
+        Note that ``quantize = True`` returns a quantized model with 8 bit
+        weights. Quantized models only support inference and run on CPUs.
+        GPU inference is not yet supported.
 
     Args:
         weights (:class:`~torchvision.models.quantization.Inception_V3_QuantizedWeights` or :class:`~torchvision.models.Inception_V3_Weights`, optional): The pretrained

torchvision/models/quantization/mobilenetv2.py

@@ -104,9 +104,10 @@ def mobilenet_v2(
     `MobileNetV2: Inverted Residuals and Linear Bottlenecks
     <https://arxiv.org/abs/1801.04381>`_.
 
-    Note that quantize = True returns a quantized model with 8 bit
-    weights. Quantized models only support inference and run on CPUs.
-    GPU inference is not yet supported
+    .. note::
+        Note that ``quantize = True`` returns a quantized model with 8 bit
+        weights. Quantized models only support inference and run on CPUs.
+        GPU inference is not yet supported.
 
     Args:
         weights (:class:`~torchvision.models.quantization.MobileNet_V2_QuantizedWeights` or :class:`~torchvision.models.MobileNet_V2_Weights`, optional): The

torchvision/models/quantization/mobilenetv3.py

@@ -200,7 +200,7 @@ def mobilenet_v3_large(
     .. note::
         Note that ``quantize = True`` returns a quantized model with 8 bit
         weights. Quantized models only support inference and run on CPUs.
-        GPU inference is not yet supported
+        GPU inference is not yet supported.
 
     Args:
         weights (:class:`~torchvision.models.quantization.MobileNet_V3_Large_QuantizedWeights` or :class:`~torchvision.models.MobileNet_V3_Large_Weights`, optional): The

torchvision/models/quantization/resnet.py

@@ -270,6 +270,11 @@ def resnet18(
     """ResNet-18 model from
     `Deep Residual Learning for Image Recognition <https://arxiv.org/abs/1512.03385.pdf>`_
 
+    .. note::
+        Note that ``quantize = True`` returns a quantized model with 8 bit
+        weights. Quantized models only support inference and run on CPUs.
+        GPU inference is not yet supported.
+
     Args:
         weights (:class:`~torchvision.models.quantization.ResNet18_QuantizedWeights` or :class:`~torchvision.models.ResNet18_Weights`, optional): The
             pretrained weights for the model. See
@@ -314,6 +319,11 @@ def resnet50(
     """ResNet-50 model from
     `Deep Residual Learning for Image Recognition <https://arxiv.org/abs/1512.03385.pdf>`_
 
+    .. note::
+        Note that ``quantize = True`` returns a quantized model with 8 bit
+        weights. Quantized models only support inference and run on CPUs.
+        GPU inference is not yet supported.
+
     Args:
         weights (:class:`~torchvision.models.quantization.ResNet50_QuantizedWeights` or :class:`~torchvision.models.ResNet50_Weights`, optional): The
             pretrained weights for the model. See
@@ -358,6 +368,11 @@ def resnext101_32x8d(
     """ResNeXt-101 32x8d model from
     `Aggregated Residual Transformation for Deep Neural Networks <https://arxiv.org/abs/1611.05431.pdf>`_
 
+    .. note::
+        Note that ``quantize = True`` returns a quantized model with 8 bit
+        weights. Quantized models only support inference and run on CPUs.
+        GPU inference is not yet supported.
+
     Args:
         weights (:class:`~torchvision.models.quantization.ResNet101_32X8D_QuantizedWeights` or :class:`~torchvision.models.ResNet101_32X8D_Weights`, optional): The
             pretrained weights for the model. See
@@ -396,6 +411,11 @@ def resnext101_64x4d(
     """ResNeXt-101 64x4d model from
     `Aggregated Residual Transformation for Deep Neural Networks <https://arxiv.org/abs/1611.05431.pdf>`_
 
+    .. note::
+        Note that ``quantize = True`` returns a quantized model with 8 bit
+        weights. Quantized models only support inference and run on CPUs.
+        GPU inference is not yet supported.
+
     Args:
         weights (:class:`~torchvision.models.quantization.ResNet101_64X4D_QuantizedWeights` or :class:`~torchvision.models.ResNet101_64X4D_Weights`, optional): The
             pretrained weights for the model. See

torchvision/models/quantization/shufflenetv2.py

@@ -208,6 +208,11 @@ def shufflenet_v2_x0_5(
     `"ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design"
     <https://arxiv.org/abs/1807.11164>`_.
 
+    .. note::
+        Note that ``quantize = True`` returns a quantized model with 8 bit
+        weights. Quantized models only support inference and run on CPUs.
+        GPU inference is not yet supported.
+
     Args:
         weights (ShuffleNet_V2_X0_5_QuantizedWeights or ShuffleNet_V2_X0_5_Weights, optional): The pretrained
             weights for the model
@@ -240,6 +245,11 @@ def shufflenet_v2_x1_0(
     `"ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design"
     <https://arxiv.org/abs/1807.11164>`_.
 
+    .. note::
+        Note that ``quantize = True`` returns a quantized model with 8 bit
+        weights. Quantized models only support inference and run on CPUs.
+        GPU inference is not yet supported.
+
     Args:
         weights (ShuffleNet_V2_X1_0_QuantizedWeights or ShuffleNet_V2_X1_0_Weights, optional): The pretrained
             weights for the model
@@ -264,6 +274,11 @@ def shufflenet_v2_x1_5(
     `"ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design"
     <https://arxiv.org/abs/1807.11164>`_.
 
+    .. note::
+        Note that ``quantize = True`` returns a quantized model with 8 bit
+        weights. Quantized models only support inference and run on CPUs.
+        GPU inference is not yet supported.
+
     Args:
         weights (ShuffleNet_V2_X1_5_QuantizedWeights or ShuffleNet_V2_X1_5_Weights, optional): The pretrained
             weights for the model
@@ -288,6 +303,11 @@ def shufflenet_v2_x2_0(
     `"ShuffleNet V2: Practical Guidelines for Efficient CNN Architecture Design"
     <https://arxiv.org/abs/1807.11164>`_.
 
+    .. note::
+        Note that ``quantize = True`` returns a quantized model with 8 bit
+        weights. Quantized models only support inference and run on CPUs.
+        GPU inference is not yet supported.
+
     Args:
         weights (ShuffleNet_V2_X2_0_QuantizedWeights or ShuffleNet_V2_X2_0_Weights, optional): The pretrained
             weights for the model