Update EfficientNet comments, MobileNetV3 non-TF create fns, fix factory arg checks, bump PyTorch version req to 1.2

Author: rwightman

requirements.txt

@@ -1,3 +1,3 @@
-torch>=1.1.0
-torchvision>=0.3.0
+torch>=1.2.0
+torchvision>=0.4.0
 pyyaml

timm/models/efficientnet.py

@@ -845,8 +845,7 @@ def spnasnet_100(pretrained=False, **kwargs):
 @register_model
 def efficientnet_b0(pretrained=False, **kwargs):
     """ EfficientNet-B0 """
-    # NOTE for train, drop_rate should be 0.2
-    #kwargs['drop_connect_rate'] = 0.2  # set when training, TODO add as cmd arg
+    # NOTE for train, drop_rate should be 0.2, drop_connect_rate should be 0.2
     model = _gen_efficientnet(
         'efficientnet_b0', channel_multiplier=1.0, depth_multiplier=1.0, pretrained=pretrained, **kwargs)
     return model
@@ -855,8 +854,7 @@ def efficientnet_b0(pretrained=False, **kwargs):
 @register_model
 def efficientnet_b1(pretrained=False, **kwargs):
     """ EfficientNet-B1 """
-    # NOTE for train, drop_rate should be 0.2
-    #kwargs['drop_connect_rate'] = 0.2  # set when training, TODO add as cmd arg
+    # NOTE for train, drop_rate should be 0.2, drop_connect_rate should be 0.2
     model = _gen_efficientnet(
         'efficientnet_b1', channel_multiplier=1.0, depth_multiplier=1.1, pretrained=pretrained, **kwargs)
     return model
@@ -865,8 +863,7 @@ def efficientnet_b1(pretrained=False, **kwargs):
 @register_model
 def efficientnet_b2(pretrained=False, **kwargs):
     """ EfficientNet-B2 """
-    # NOTE for train, drop_rate should be 0.3
-    #kwargs['drop_connect_rate'] = 0.2  # set when training, TODO add as cmd arg
+    # NOTE for train, drop_rate should be 0.3, drop_connect_rate should be 0.2
     model = _gen_efficientnet(
         'efficientnet_b2', channel_multiplier=1.1, depth_multiplier=1.2, pretrained=pretrained, **kwargs)
     return model
@@ -875,8 +872,7 @@ def efficientnet_b2(pretrained=False, **kwargs):
 @register_model
 def efficientnet_b3(pretrained=False, **kwargs):
     """ EfficientNet-B3 """
-    # NOTE for train, drop_rate should be 0.3
-    #kwargs['drop_connect_rate'] = 0.2  # set when training, TODO add as cmd arg
+    # NOTE for train, drop_rate should be 0.3, drop_connect_rate should be 0.2
     model = _gen_efficientnet(
         'efficientnet_b3', channel_multiplier=1.2, depth_multiplier=1.4, pretrained=pretrained, **kwargs)
     return model
@@ -885,8 +881,7 @@ def efficientnet_b3(pretrained=False, **kwargs):
 @register_model
 def efficientnet_b4(pretrained=False, **kwargs):
     """ EfficientNet-B4 """
-    # NOTE for train, drop_rate should be 0.4
-    #kwargs['drop_connect_rate'] = 0.2  #  set when training, TODO add as cmd arg
+    # NOTE for train, drop_rate should be 0.4, drop_connect_rate should be 0.2
     model = _gen_efficientnet(
         'efficientnet_b4', channel_multiplier=1.4, depth_multiplier=1.8, pretrained=pretrained, **kwargs)
     return model
@@ -895,8 +890,7 @@ def efficientnet_b4(pretrained=False, **kwargs):
 @register_model
 def efficientnet_b5(pretrained=False, **kwargs):
     """ EfficientNet-B5 """
-    # NOTE for train, drop_rate should be 0.4
-    #kwargs['drop_connect_rate'] = 0.2  # set when training, TODO add as cmd arg
+    # NOTE for train, drop_rate should be 0.4, drop_connect_rate should be 0.2
     model = _gen_efficientnet(
         'efficientnet_b5', channel_multiplier=1.6, depth_multiplier=2.2, pretrained=pretrained, **kwargs)
     return model
@@ -905,8 +899,7 @@ def efficientnet_b5(pretrained=False, **kwargs):
 @register_model
 def efficientnet_b6(pretrained=False, **kwargs):
     """ EfficientNet-B6 """
-    # NOTE for train, drop_rate should be 0.5
-    #kwargs['drop_connect_rate'] = 0.2  # set when training, TODO add as cmd arg
+    # NOTE for train, drop_rate should be 0.5, drop_connect_rate should be 0.2
     model = _gen_efficientnet(
         'efficientnet_b6', channel_multiplier=1.8, depth_multiplier=2.6, pretrained=pretrained, **kwargs)
     return model
@@ -915,8 +908,7 @@ def efficientnet_b6(pretrained=False, **kwargs):
 @register_model
 def efficientnet_b7(pretrained=False, **kwargs):
     """ EfficientNet-B7 """
-    # NOTE for train, drop_rate should be 0.5
-    #kwargs['drop_connect_rate'] = 0.2  # set when training, TODO add as cmd arg
+    # NOTE for train, drop_rate should be 0.5, drop_connect_rate should be 0.2
     model = _gen_efficientnet(
         'efficientnet_b7', channel_multiplier=2.0, depth_multiplier=3.1, pretrained=pretrained, **kwargs)
     return model
@@ -949,8 +941,7 @@ def efficientnet_el(pretrained=False, **kwargs):
 @register_model
 def efficientnet_cc_b0_4e(pretrained=False, **kwargs):
     """ EfficientNet-CondConv-B0 w/ 8 Experts """
-    # NOTE for train, drop_rate should be 0.2
-    #kwargs['drop_connect_rate'] = 0.2  # set when training, TODO add as cmd arg
+    # NOTE for train, drop_rate should be 0.2, drop_connect_rate should be 0.2
     model = _gen_efficientnet_condconv(
         'efficientnet_cc_b0_4e', channel_multiplier=1.0, depth_multiplier=1.0, pretrained=pretrained, **kwargs)
     return model
@@ -959,8 +950,7 @@ def efficientnet_cc_b0_4e(pretrained=False, **kwargs):
 @register_model
 def efficientnet_cc_b0_8e(pretrained=False, **kwargs):
     """ EfficientNet-CondConv-B0 w/ 8 Experts """
-    # NOTE for train, drop_rate should be 0.2
-    #kwargs['drop_connect_rate'] = 0.2  # set when training, TODO add as cmd arg
+    # NOTE for train, drop_rate should be 0.2, drop_connect_rate should be 0.2
     model = _gen_efficientnet_condconv(
         'efficientnet_cc_b0_8e', channel_multiplier=1.0, depth_multiplier=1.0, experts_multiplier=2,
         pretrained=pretrained, **kwargs)
@@ -969,8 +959,7 @@ def efficientnet_cc_b0_8e(pretrained=False, **kwargs):
 @register_model
 def efficientnet_cc_b1_8e(pretrained=False, **kwargs):
     """ EfficientNet-CondConv-B1 w/ 8 Experts """
-    # NOTE for train, drop_rate should be 0.2
-    #kwargs['drop_connect_rate'] = 0.2  # set when training, TODO add as cmd arg
+    # NOTE for train, drop_rate should be 0.2, drop_connect_rate should be 0.2
     model = _gen_efficientnet_condconv(
         'efficientnet_cc_b1_8e', channel_multiplier=1.0, depth_multiplier=1.1, experts_multiplier=2,
         pretrained=pretrained, **kwargs)
@@ -1008,7 +997,7 @@ def tf_efficientnet_b2(pretrained=False, **kwargs):
 
 
 @register_model
-def tf_efficientnet_b3(pretrained=False, num_classes=1000, in_chans=3, **kwargs):
+def tf_efficientnet_b3(pretrained=False, **kwargs):
     """ EfficientNet-B3. Tensorflow compatible variant """
     kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
     kwargs['pad_type'] = 'same'
@@ -1090,7 +1079,7 @@ def tf_efficientnet_b2_ap(pretrained=False, **kwargs):
 
 
 @register_model
-def tf_efficientnet_b3_ap(pretrained=False, num_classes=1000, in_chans=3, **kwargs):
+def tf_efficientnet_b3_ap(pretrained=False, **kwargs):
     """ EfficientNet-B3. Tensorflow compatible variant """
     kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
     kwargs['pad_type'] = 'same'
@@ -1186,8 +1175,7 @@ def tf_efficientnet_el(pretrained=False, **kwargs):
 @register_model
 def tf_efficientnet_cc_b0_4e(pretrained=False, **kwargs):
     """ EfficientNet-CondConv-B0 w/ 4 Experts. Tensorflow compatible variant """
-    # NOTE for train, drop_rate should be 0.2
-    #kwargs['drop_connect_rate'] = 0.2  # set when training, TODO add as cmd arg
+    # NOTE for train, drop_rate should be 0.2, drop_connect_rate should be 0.2
     kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
     kwargs['pad_type'] = 'same'
     model = _gen_efficientnet_condconv(
@@ -1198,8 +1186,7 @@ def tf_efficientnet_cc_b0_4e(pretrained=False, **kwargs):
 @register_model
 def tf_efficientnet_cc_b0_8e(pretrained=False, **kwargs):
     """ EfficientNet-CondConv-B0 w/ 8 Experts. Tensorflow compatible variant """
-    # NOTE for train, drop_rate should be 0.2
-    #kwargs['drop_connect_rate'] = 0.2  # set when training, TODO add as cmd arg
+    # NOTE for train, drop_rate should be 0.2, drop_connect_rate should be 0.2
     kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
     kwargs['pad_type'] = 'same'
     model = _gen_efficientnet_condconv(
@@ -1210,8 +1197,7 @@ def tf_efficientnet_cc_b0_8e(pretrained=False, **kwargs):
 @register_model
 def tf_efficientnet_cc_b1_8e(pretrained=False, **kwargs):
     """ EfficientNet-CondConv-B1 w/ 8 Experts. Tensorflow compatible variant """
-    # NOTE for train, drop_rate should be 0.2
-    #kwargs['drop_connect_rate'] = 0.2  # set when training, TODO add as cmd arg
+    # NOTE for train, drop_rate should be 0.2, drop_connect_rate should be 0.2
     kwargs['bn_eps'] = BN_EPS_TF_DEFAULT
     kwargs['pad_type'] = 'same'
     model = _gen_efficientnet_condconv(
@@ -1262,7 +1248,6 @@ def mixnet_xxl(pretrained=False, **kwargs):
     """Creates a MixNet Double Extra Large model.
     Not a paper spec, experimental def by RW w/ depth scaling.
     """
-    # kwargs['drop_connect_rate'] = 0.2
     model = _gen_mixnet_m(
         'mixnet_xxl', channel_multiplier=2.4, depth_multiplier=1.3, pretrained=pretrained, **kwargs)
     return model

timm/models/factory.py

@@ -25,8 +25,8 @@ def create_model(
     """
     margs = dict(pretrained=pretrained, num_classes=num_classes, in_chans=in_chans)
 
-    # Only gen_efficientnet models have support for batchnorm params or drop_connect_rate passed as args
-    is_efficientnet = is_model_in_modules(model_name, ['gen_efficientnet'])
+    # Only EfficientNet and MobileNetV3 models have support for batchnorm params or drop_connect_rate passed as args
+    is_efficientnet = is_model_in_modules(model_name, ['efficientnet', 'mobilenetv3'])
     if not is_efficientnet:
         kwargs.pop('bn_tf', None)
         kwargs.pop('bn_momentum', None)

timm/models/mobilenetv3.py

@@ -35,6 +35,8 @@ def _cfg(url='', **kwargs):
 default_cfgs = {
     'mobilenetv3_large_075': _cfg(url=''),
     'mobilenetv3_large_100': _cfg(url=''),
+    'mobilenetv3_small_075': _cfg(url=''),
+    'mobilenetv3_small_100': _cfg(url=''),
     'mobilenetv3_rw': _cfg(
         url='https://github.com/rwightman/pytorch-image-models/releases/download/v0.1-weights/mobilenetv3_100-35495452.pth',
         interpolation='bicubic'),
@@ -374,6 +376,35 @@ def _gen_mobilenet_v3(variant, channel_multiplier=1.0, pretrained=False, **kwarg
     return model
 
 
+@register_model
+def mobilenetv3_large_075(pretrained=False, **kwargs):
+    """ MobileNet V3 """
+    model = _gen_mobilenet_v3('mobilenetv3_large_075', 0.75, pretrained=pretrained, **kwargs)
+    return model
+
+
+@register_model
+def mobilenetv3_large_100(pretrained=False, **kwargs):
+    """ MobileNet V3 """
+    model = _gen_mobilenet_v3('mobilenetv3_large_100', 1.0, pretrained=pretrained, **kwargs)
+    return model
+
+
+@register_model
+def mobilenetv3_small_075(pretrained=False, **kwargs):
+    """ MobileNet V3 """
+    model = _gen_mobilenet_v3('mobilenetv3_small_075', 0.75, pretrained=pretrained, **kwargs)
+    return model
+
+
+@register_model
+def mobilenetv3_small_100(pretrained=False, **kwargs):
+    print(kwargs)
+    """ MobileNet V3 """
+    model = _gen_mobilenet_v3('mobilenetv3_small_100', 1.0, pretrained=pretrained, **kwargs)
+    return model
+
+
 @register_model
 def mobilenetv3_rw(pretrained=False, **kwargs):
     """ MobileNet V3 """
@@ -384,7 +415,6 @@ def mobilenetv3_rw(pretrained=False, **kwargs):
     return model
 
 
-
 @register_model
 def tf_mobilenetv3_large_075(pretrained=False, **kwargs):
     """ MobileNet V3 """