Cleanup model_factory imports, consistent __all__ for models, fixed inception_v4 weight url

Author: rwightman

models/densenet.py

@@ -12,7 +12,8 @@
 from data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
 import re
 
-__all__ = ['DenseNet', 'densenet121', 'densenet169', 'densenet201', 'densenet161']
+_models = ['densenet121', 'densenet169', 'densenet201', 'densenet161']
+__all__ = ['DenseNet'] + _models
 
 
 def _cfg(url=''):

models/dpn.py

@@ -19,7 +19,8 @@
 from models.adaptive_avgmax_pool import select_adaptive_pool2d
 from data import IMAGENET_DPN_MEAN, IMAGENET_DPN_STD
 
-__all__ = ['DPN', 'dpn68', 'dpn92', 'dpn98', 'dpn131', 'dpn107']
+_models = ['dpn68', 'dpn68b', 'dpn92', 'dpn98', 'dpn131', 'dpn107']
+__all__ = ['DPN'] + _models
 
 
 def _cfg(url=''):
@@ -32,18 +33,12 @@ def _cfg(url=''):
 
 
 default_cfgs = {
-    'dpn68':
-        _cfg(url='http://data.lip6.fr/cadene/pretrainedmodels/dpn68-66bebafa7.pth'),
-    'dpn68b_extra':
-        _cfg(url='http://data.lip6.fr/cadene/pretrainedmodels/dpn68b_extra-84854c156.pth'),
-    'dpn92_extra':
-        _cfg(url='http://data.lip6.fr/cadene/pretrainedmodels/dpn92_extra-b040e4a9b.pth'),
-    'dpn98':
-        _cfg(url='http://data.lip6.fr/cadene/pretrainedmodels/dpn98-5b90dec4d.pth'),
-    'dpn131':
-        _cfg(url='http://data.lip6.fr/cadene/pretrainedmodels/dpn131-71dfe43e0.pth'),
-    'dpn107_extra':
-        _cfg(url='http://data.lip6.fr/cadene/pretrainedmodels/dpn107_extra-1ac7121e2.pth')
+    'dpn68': _cfg(url='http://data.lip6.fr/cadene/pretrainedmodels/dpn68-66bebafa7.pth'),
+    'dpn68b_extra': _cfg(url='http://data.lip6.fr/cadene/pretrainedmodels/dpn68b_extra-84854c156.pth'),
+    'dpn92_extra': _cfg(url='http://data.lip6.fr/cadene/pretrainedmodels/dpn92_extra-b040e4a9b.pth'),
+    'dpn98': _cfg(url='http://data.lip6.fr/cadene/pretrainedmodels/dpn98-5b90dec4d.pth'),
+    'dpn131': _cfg(url='http://data.lip6.fr/cadene/pretrainedmodels/dpn131-71dfe43e0.pth'),
+    'dpn107_extra': _cfg(url='http://data.lip6.fr/cadene/pretrainedmodels/dpn107_extra-1ac7121e2.pth')
 }
 
 

models/genmobilenet.py

@@ -26,10 +26,12 @@
 from models.conv2d_same import sconv2d
 from data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
 
-__all__ = ['GenMobileNet', 'mnasnet_050', 'mnasnet_075', 'mnasnet_100', 'mnasnet_140',
-           'semnasnet_050', 'semnasnet_075', 'semnasnet_100', 'semnasnet_140', 'mnasnet_small',
-           'mobilenetv1_100', 'mobilenetv2_100', 'mobilenetv3_050', 'mobilenetv3_075', 'mobilenetv3_100',
-           'chamnetv1_100', 'chamnetv2_100', 'fbnetc_100', 'spnasnet_100']
+_models = [
+    'mnasnet_050', 'mnasnet_075', 'mnasnet_100', 'mnasnet_140', 'semnasnet_050', 'semnasnet_075',
+    'semnasnet_100', 'semnasnet_140', 'mnasnet_small', 'mobilenetv1_100', 'mobilenetv2_100',
+    'mobilenetv3_050', 'mobilenetv3_075', 'mobilenetv3_100', 'chamnetv1_100', 'chamnetv2_100',
+    'fbnetc_100', 'spnasnet_100', 'tflite_mnasnet_100', 'tflite_semnasnet_100']
+__all__ = ['GenMobileNet', 'genmobilenet_model_names'] + _models
 
 
 def _cfg(url='', **kwargs):
@@ -67,7 +69,7 @@ def _cfg(url='', **kwargs):
     'spnasnet_100': _cfg(url='https://www.dropbox.com/s/iieopt18rytkgaa/spnasnet_100-048bc3f4.pth?dl=1'),
 }
 
-_DEBUG = True
+_DEBUG = False
 
 # Default args for PyTorch BN impl
 _BN_MOMENTUM_PT_DEFAULT = 0.1
@@ -266,7 +268,7 @@ class _BlockBuilder:
     def __init__(self, depth_multiplier=1.0, depth_divisor=8, min_depth=None,
                  act_fn=None, se_gate_fn=torch.sigmoid, se_reduce_mid=False,
                  bn_momentum=_BN_MOMENTUM_PT_DEFAULT, bn_eps=_BN_EPS_PT_DEFAULT,
-                 folded_bn=False, padding_same=False):
+                 folded_bn=False, padding_same=False, verbose=False):
         self.depth_multiplier = depth_multiplier
         self.depth_divisor = depth_divisor
         self.min_depth = min_depth
@@ -277,6 +279,7 @@ def __init__(self, depth_multiplier=1.0, depth_divisor=8, min_depth=None,
         self.bn_eps = bn_eps
         self.folded_bn = folded_bn
         self.padding_same = padding_same
+        self.verbose = verbose
         self.in_chs = None
 
     def _round_channels(self, chs):
@@ -293,7 +296,7 @@ def _make_block(self, ba):
         # block act fn overrides the model default
         ba['act_fn'] = ba['act_fn'] if ba['act_fn'] is not None else self.act_fn
         assert ba['act_fn'] is not None
-        if _DEBUG:
+        if self.verbose:
             print('args:', ba)
         #  could replace this if with lambdas or functools binding if variety increases
         if bt == 'ir':
@@ -315,7 +318,7 @@ def _make_stack(self, stack_args):
         blocks = []
         # each stack (stage) contains a list of block arguments
         for block_idx, ba in enumerate(stack_args):
-            if _DEBUG:
+            if self.verbose:
                 print('block', block_idx, end=', ')
             if block_idx >= 1:
                 # only the first block in any stack/stage can have a stride > 1
@@ -334,18 +337,18 @@ def __call__(self, in_chs, arch_def):
              List of block stacks (each stack wrapped in nn.Sequential)
         """
         arch_args = _decode_arch_def(arch_def)  # convert and expand string defs to arg dicts
-        if _DEBUG:
+        if self.verbose:
             print('Building model trunk with %d stacks (stages)...' % len(arch_args))
         self.in_chs = in_chs
         blocks = []
         # outer list of arch_args defines the stacks ('stages' by some conventions)
         for stack_idx, stack in enumerate(arch_args):
-            if _DEBUG:
+            if self.verbose:
                 print('stack', stack_idx)
             assert isinstance(stack, list)
             stack = self._make_stack(stack)
             blocks.append(stack)
-            if _DEBUG:
+            if self.verbose:
                 print()
         return blocks
 
@@ -631,7 +634,7 @@ def __init__(self, block_args, num_classes=1000, in_chans=3, stem_size=32, num_f
         builder = _BlockBuilder(
             depth_multiplier, depth_divisor, min_depth,
             act_fn, se_gate_fn, se_reduce_mid,
-            bn_momentum, bn_eps, folded_bn, padding_same)
+            bn_momentum, bn_eps, folded_bn, padding_same, verbose=_DEBUG)
         self.blocks = nn.Sequential(*builder(in_chs, block_args))
         in_chs = builder.in_chs
 
@@ -1265,3 +1268,7 @@ def spnasnet_100(num_classes, in_chans=3, pretrained=False, **kwargs):
     if pretrained:
         load_pretrained(model, default_cfg, num_classes, in_chans)
     return model
+
+
+def genmobilenet_model_names():
+    return set(_models)

models/gluon_resnet.py

@@ -11,13 +11,14 @@
 from models.adaptive_avgmax_pool import SelectAdaptivePool2d
 from data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
 
-__all__ = ['GluonResNet', 'gluon_resnet18_v1b', 'gluon_resnet34_v1b', 'gluon_resnet50_v1b', 'gluon_resnet101_v1b',
-    'gluon_resnet152_v1b', 'gluon_resnet50_v1c', 'gluon_resnet101_v1c', 'gluon_resnet152_v1c', 'gluon_resnet50_v1d',
-    'gluon_resnet101_v1d', 'gluon_resnet152_v1d', 'gluon_resnet50_v1e', 'gluon_resnet101_v1e', 'gluon_resnet152_v1e',
-    'gluon_resnet50_v1s', 'gluon_resnet101_v1s', 'gluon_resnet152_v1s', 'gluon_resnext50_32x4d',
-    'gluon_resnext101_32x4d', 'gluon_resnext101_64x4d', 'gluon_resnext152_32x4d', 'gluon_seresnext50_32x4d',
-    'gluon_seresnext101_32x4d', 'gluon_seresnext101_64x4d', 'gluon_seresnext152_32x4d', 'gluon_senet154'
-]
+_models = [
+    'gluon_resnet18_v1b', 'gluon_resnet34_v1b', 'gluon_resnet50_v1b', 'gluon_resnet101_v1b', 'gluon_resnet152_v1b',
+    'gluon_resnet50_v1c', 'gluon_resnet101_v1c', 'gluon_resnet152_v1c', 'gluon_resnet50_v1d', 'gluon_resnet101_v1d',
+    'gluon_resnet152_v1d', 'gluon_resnet50_v1e', 'gluon_resnet101_v1e', 'gluon_resnet152_v1e', 'gluon_resnet50_v1s',
+    'gluon_resnet101_v1s', 'gluon_resnet152_v1s', 'gluon_resnext50_32x4d', 'gluon_resnext101_32x4d',
+    'gluon_resnext101_64x4d', 'gluon_resnext152_32x4d', 'gluon_seresnext50_32x4d', 'gluon_seresnext101_32x4d',
+    'gluon_seresnext101_64x4d', 'gluon_seresnext152_32x4d', 'gluon_senet154']
+__all__ = ['GluonResNet'] + _models
 
 
 def _cfg(url='', **kwargs):

models/inception_resnet_v2.py

@@ -9,6 +9,9 @@
 from models.adaptive_avgmax_pool import *
 from data import IMAGENET_INCEPTION_MEAN, IMAGENET_INCEPTION_STD
 
+_models = ['inception_resnet_v2']
+__all__ = ['InceptionResnetV2'] + _models
+
 default_cfgs = {
     'inception_resnet_v2': {
         'url': 'http://data.lip6.fr/cadene/pretrainedmodels/inceptionresnetv2-520b38e4.pth',

models/inception_v3.py

@@ -2,6 +2,9 @@
 from models.helpers import load_pretrained
 from data import IMAGENET_DEFAULT_STD, IMAGENET_DEFAULT_MEAN, IMAGENET_INCEPTION_MEAN, IMAGENET_INCEPTION_STD
 
+_models = ['inception_v3', 'tf_inception_v3', 'adv_inception_v3', 'gluon_inception_v3']
+__all__ = _models
+
 default_cfgs = {
     # original PyTorch weights, ported from Tensorflow but modified
     'inception_v3': {

models/inception_v4.py

@@ -9,13 +9,16 @@
 from models.adaptive_avgmax_pool import *
 from data import IMAGENET_INCEPTION_MEAN, IMAGENET_INCEPTION_STD
 
+_models = ['inception_v4']
+__all__ = ['InceptionV4'] + _models
+
 default_cfgs = {
     'inception_v4': {
-        'url': 'http://webia.lip6.fr/~cadene/Downloads/inceptionv4-97ef9c30.pth',
+        'url': 'http://data.lip6.fr/cadene/pretrainedmodels/inceptionv4-8e4777a0.pth',
         'num_classes': 1001, 'input_size': (3, 299, 299), 'pool_size': (8, 8),
         'crop_pct': 0.875, 'interpolation': 'bicubic',
         'mean': IMAGENET_INCEPTION_MEAN, 'std': IMAGENET_INCEPTION_STD,
-        'first_conv': 'features.0.conv', 'classifier': 'classif',
+        'first_conv': 'features.0.conv', 'classifier': 'last_linear',
     }
 }
 
@@ -268,7 +271,7 @@ def __init__(self, num_classes=1001, in_chans=3, drop_rate=0., global_pool='avg'
             Inception_C(),
             Inception_C(),
         )
-        self.classif = nn.Linear(self.num_features, num_classes)
+        self.last_linear = nn.Linear(self.num_features, num_classes)
 
     def get_classifier(self):
         return self.classif
@@ -289,7 +292,7 @@ def forward(self, x):
         x = self.forward_features(x)
         if self.drop_rate > 0:
             x = F.dropout(x, p=self.drop_rate, training=self.training)
-        x = self.classif(x)
+        x = self.last_linear(x)
         return x
 
 

models/model_factory.py

@@ -1,38 +1,18 @@
-from models.inception_v4 import inception_v4
-from models.inception_resnet_v2 import inception_resnet_v2
-from models.densenet import densenet161, densenet121, densenet169, densenet201
-from models.resnet import resnet18, resnet34, resnet50, resnet101, resnet152, \
-    resnext50_32x4d, resnext101_32x4d, resnext101_64x4d, resnext152_32x4d
-from models.dpn import dpn68, dpn68b, dpn92, dpn98, dpn131, dpn107
-from models.senet import seresnet18, seresnet34, seresnet50, seresnet101, seresnet152, \
-    seresnext26_32x4d, seresnext50_32x4d, seresnext101_32x4d
-from models.xception import xception
-from models.pnasnet import pnasnet5large
-from models.genmobilenet import \
-    mnasnet_050, mnasnet_075, mnasnet_100, mnasnet_140, tflite_mnasnet_100,\
-    semnasnet_050, semnasnet_075, semnasnet_100, semnasnet_140, tflite_semnasnet_100, mnasnet_small,\
-    mobilenetv1_100, mobilenetv2_100, mobilenetv3_050, mobilenetv3_075, mobilenetv3_100,\
-    fbnetc_100, chamnetv1_100, chamnetv2_100, spnasnet_100
-from models.inception_v3 import inception_v3, gluon_inception_v3, tf_inception_v3, adv_inception_v3
-from models.gluon_resnet import gluon_resnet18_v1b, gluon_resnet34_v1b, gluon_resnet50_v1b, gluon_resnet101_v1b, \
-    gluon_resnet152_v1b, gluon_resnet50_v1c, gluon_resnet101_v1c, gluon_resnet152_v1c, \
-    gluon_resnet50_v1d, gluon_resnet101_v1d, gluon_resnet152_v1d, \
-    gluon_resnet50_v1e, gluon_resnet101_v1e, gluon_resnet152_v1e, \
-    gluon_resnet50_v1s, gluon_resnet101_v1s, gluon_resnet152_v1s, \
-    gluon_resnext50_32x4d, gluon_resnext101_32x4d , gluon_resnext101_64x4d, gluon_resnext152_32x4d, \
-    gluon_seresnext50_32x4d, gluon_seresnext101_32x4d, gluon_seresnext101_64x4d, gluon_seresnext152_32x4d, \
-    gluon_senet154
+from models.inception_v4 import *
+from models.inception_resnet_v2 import *
+from models.densenet import *
+from models.resnet import *
+from models.dpn import *
+from models.senet import *
+from models.xception import *
+from models.pnasnet import *
+from models.genmobilenet import *
+from models.inception_v3 import *
+from models.gluon_resnet import *
 
 from models.helpers import load_checkpoint
 
 
-def _is_genmobilenet(name):
-    genmobilenets = ['mnasnet', 'semnasnet', 'fbnet', 'chamnet', 'mobilenet']
-    if any([name.startswith(x) for x in genmobilenets]):
-        return True
-    return False
-
-
 def create_model(
         model_name='resnet50',
         pretrained=None,
@@ -44,8 +24,7 @@ def create_model(
     margs = dict(num_classes=num_classes, in_chans=in_chans, pretrained=pretrained)
 
     # Not all models have support for batchnorm params passed as args, only genmobilenet variants
-    # FIXME better way to do this without pushing support into every other model fn?
-    supports_bn_params = _is_genmobilenet(model_name)
+    supports_bn_params = model_name in genmobilenet_model_names()
     if not supports_bn_params and any([x in kwargs for x in ['bn_tf', 'bn_momentum', 'bn_eps']]):
         kwargs.pop('bn_tf', None)
         kwargs.pop('bn_momentum', None)

models/pnasnet.py

@@ -15,6 +15,9 @@
 from models.helpers import load_pretrained
 from models.adaptive_avgmax_pool import SelectAdaptivePool2d
 
+_models = ['pnasnet5large']
+__all__ = ['PNASNet5Large'] + _models
+
 default_cfgs = {
     'pnasnet5large': {
         'url': 'http://data.lip6.fr/cadene/pretrainedmodels/pnasnet5large-bf079911.pth',

models/resnet.py

@@ -12,8 +12,9 @@
 from models.adaptive_avgmax_pool import SelectAdaptivePool2d
 from data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD
 
-__all__ = ['ResNet', 'resnet18', 'resnet34', 'resnet50', 'resnet101', 'resnet152',
+_models = ['resnet18', 'resnet34', 'resnet50', 'resnet101', 'resnet152',
            'resnext50_32x4d', 'resnext101_32x4d', 'resnext101_64x4d', 'resnext152_32x4d']
+__all__ = ['ResNet'] + _models
 
 
 def _cfg(url='', **kwargs):