correct pytorch model sub-architecture names (#57)

Author: bfineran

src/sparseml/pytorch/models/classification/inception_v3.py

@@ -470,7 +470,7 @@ def forward(self, x_tens: Tensor) -> Tuple[Tensor, ...]:
     domain="cv",
     sub_domain="classification",
     architecture="inception_v3",
-    sub_architecture="none",
+    sub_architecture=None,
     default_dataset="imagenet",
     default_desc="base",
     def_ignore_error_tensors=[

src/sparseml/pytorch/models/classification/resnet.py

@@ -825,7 +825,7 @@ def resnetv2_50(num_classes: int = 1000, class_type: str = "single") -> ResNet:
     domain="cv",
     sub_domain="classification",
     architecture="resnet_v1",
-    sub_architecture="50-2xwidth",
+    sub_architecture="50_2x",
     default_dataset="imagenet",
     default_desc="base",
     def_ignore_error_tensors=["classifier.fc.weight", "classifier.fc.bias"],
@@ -1067,7 +1067,7 @@ def resnetv2_101(num_classes: int = 1000, class_type: str = "single") -> ResNet:
     domain="cv",
     sub_domain="classification",
     architecture="resnet_v1",
-    sub_architecture="101-2xwidth",
+    sub_architecture="101_2x",
     default_dataset="imagenet",
     default_desc="base",
     def_ignore_error_tensors=["classifier.fc.weight", "classifier.fc.bias"],

src/sparseml/pytorch/models/classification/vgg.py

@@ -238,7 +238,7 @@ def vgg11(num_classes: int = 1000, class_type: str = "single") -> VGG:
     domain="cv",
     sub_domain="classification",
     architecture="vgg",
-    sub_architecture="11-bn",
+    sub_architecture="11_bn",
     default_dataset="imagenet",
     default_desc="base",
     def_ignore_error_tensors=["classifier.mlp.6.weight", "classifier.mlp.6.bias"],
@@ -324,7 +324,7 @@ def vgg13(num_classes: int = 1000, class_type: str = "single") -> VGG:
     domain="cv",
     sub_domain="classification",
     architecture="vgg",
-    sub_architecture="13-bn",
+    sub_architecture="13_bn",
     default_dataset="imagenet",
     default_desc="base",
     def_ignore_error_tensors=["classifier.mlp.6.weight", "classifier.mlp.6.bias"],
@@ -410,7 +410,7 @@ def vgg16(num_classes: int = 1000, class_type: str = "single") -> VGG:
     domain="cv",
     sub_domain="classification",
     architecture="vgg",
-    sub_architecture="16-bn",
+    sub_architecture="16_bn",
     default_dataset="imagenet",
     default_desc="base",
     def_ignore_error_tensors=["classifier.mlp.6.weight", "classifier.mlp.6.bias"],
@@ -496,7 +496,7 @@ def vgg19(num_classes: int = 1000, class_type: str = "single") -> VGG:
     domain="cv",
     sub_domain="classification",
     architecture="vgg",
-    sub_architecture="19-bn",
+    sub_architecture="19_bn",
     default_dataset="imagenet",
     default_desc="base",
     def_ignore_error_tensors=["classifier.mlp.6.weight", "classifier.mlp.6.bias"],

src/sparseml/pytorch/models/detection/yolo_v3.py

@@ -276,7 +276,7 @@ def forward(self, inp: Tensor):
     domain="cv",
     sub_domain="detection",
     architecture="yolo_v3",
-    sub_architecture="none",
+    sub_architecture="spp",
     default_dataset="coco",
     default_desc="base",
 )

src/sparseml/tensorflow_v1/optim/mask_pruning.py

@@ -20,8 +20,10 @@
 from collections import namedtuple
 from typing import List, Tuple
 
+
 try:
     import tensorflow.contrib.graph_editor as graph_editor
+
     tf_contrib_err = None
 except Exception as err:
     graph_editor = None

src/sparseml/tensorflow_v1/utils/variable.py

@@ -17,6 +17,7 @@
 
 import numpy
 
+
 try:
     import tensorflow.contrib.graph_editor as graph_editor
     from tensorflow.contrib.graph_editor.util import ListView
@@ -239,7 +240,9 @@ def get_ops_and_inputs_by_name_or_regex(
                         nm_ks_consuming_ops_with_input = [
                             (consuming_op, inp)
                             for output_tens in graph_editor.sgv(op).outputs
-                            for consuming_op in graph_editor.get_consuming_ops(output_tens)
+                            for consuming_op in graph_editor.get_consuming_ops(
+                                output_tens
+                            )
                             if "_nm_ks" not in consuming_op.name
                         ]
                         prunable_ops_and_inputs += nm_ks_consuming_ops_with_input