Cleanup weight docs (#7074)

* _weight_size -> _file_size

* Better formatting of individual weights tables

* Remove file size from main tables to avoid confusion with weight size (as in RAM)

* Remove unnecessary (file size) suffix

* Fix CI error?

* Formatting
Co-authored-by: Philip Meier <github.pmeier@posteo.de>

torchvision/models/vision_transformer.py

@@ -364,7 +364,7 @@ class ViT_B_16_Weights(WeightsEnum):
                 }
             },
             "_ops": 17.564,
-            "_weight_size": 330.285,
+            "_file_size": 330.285,
             "_docs": """
                 These weights were trained from scratch by using a modified version of `DeIT
                 <https://arxiv.org/abs/2012.12877>`_'s training recipe.
@@ -390,7 +390,7 @@ class ViT_B_16_Weights(WeightsEnum):
                 }
             },
             "_ops": 55.484,
-            "_weight_size": 331.398,
+            "_file_size": 331.398,
             "_docs": """
                 These weights are learnt via transfer learning by end-to-end fine-tuning the original
                 `SWAG <https://arxiv.org/abs/2201.08371>`_ weights on ImageNet-1K data.
@@ -417,7 +417,7 @@ class ViT_B_16_Weights(WeightsEnum):
                 }
             },
             "_ops": 17.564,
-            "_weight_size": 330.285,
+            "_file_size": 330.285,
             "_docs": """
                 These weights are composed of the original frozen `SWAG <https://arxiv.org/abs/2201.08371>`_ trunk
                 weights and a linear classifier learnt on top of them trained on ImageNet-1K data.
@@ -443,7 +443,7 @@ class ViT_B_32_Weights(WeightsEnum):
                 }
             },
             "_ops": 4.409,
-            "_weight_size": 336.604,
+            "_file_size": 336.604,
             "_docs": """
                 These weights were trained from scratch by using a modified version of `DeIT
                 <https://arxiv.org/abs/2012.12877>`_'s training recipe.
@@ -469,7 +469,7 @@ class ViT_L_16_Weights(WeightsEnum):
                 }
             },
             "_ops": 61.555,
-            "_weight_size": 1161.023,
+            "_file_size": 1161.023,
             "_docs": """
                 These weights were trained from scratch by using a modified version of TorchVision's
                 `new training recipe
@@ -496,7 +496,7 @@ class ViT_L_16_Weights(WeightsEnum):
                 }
             },
             "_ops": 361.986,
-            "_weight_size": 1164.258,
+            "_file_size": 1164.258,
             "_docs": """
                 These weights are learnt via transfer learning by end-to-end fine-tuning the original
                 `SWAG <https://arxiv.org/abs/2201.08371>`_ weights on ImageNet-1K data.
@@ -523,7 +523,7 @@ class ViT_L_16_Weights(WeightsEnum):
                 }
             },
             "_ops": 61.555,
-            "_weight_size": 1161.023,
+            "_file_size": 1161.023,
             "_docs": """
                 These weights are composed of the original frozen `SWAG <https://arxiv.org/abs/2201.08371>`_ trunk
                 weights and a linear classifier learnt on top of them trained on ImageNet-1K data.
@@ -549,7 +549,7 @@ class ViT_L_32_Weights(WeightsEnum):
                 }
             },
             "_ops": 15.378,
-            "_weight_size": 1169.449,
+            "_file_size": 1169.449,
             "_docs": """
                 These weights were trained from scratch by using a modified version of `DeIT
                 <https://arxiv.org/abs/2012.12877>`_'s training recipe.
@@ -579,7 +579,7 @@ class ViT_H_14_Weights(WeightsEnum):
                 }
             },
             "_ops": 1016.717,
-            "_weight_size": 2416.643,
+            "_file_size": 2416.643,
             "_docs": """
                 These weights are learnt via transfer learning by end-to-end fine-tuning the original
                 `SWAG <https://arxiv.org/abs/2201.08371>`_ weights on ImageNet-1K data.
@@ -606,7 +606,7 @@ class ViT_H_14_Weights(WeightsEnum):
                 }
             },
             "_ops": 167.295,
-            "_weight_size": 2411.209,
+            "_file_size": 2411.209,
             "_docs": """
                 These weights are composed of the original frozen `SWAG <https://arxiv.org/abs/2201.08371>`_ trunk
                 weights and a linear classifier learnt on top of them trained on ImageNet-1K data.