From 2ac663f340e2be4d3da96c24f2c23a5e9f01c91d Mon Sep 17 00:00:00 2001 From: Ross Wightman Date: Tue, 21 Jul 2020 11:15:30 -0700 Subject: [PATCH] Add feature support to legacy senets, add 32x32 resnext models to exclude list for feature testing. --- tests/test_models.py | 4 +- timm/models/senet.py | 160 ++++++++++++-------------------- timm/models/xception_aligned.py | 5 +- 3 files changed, 65 insertions(+), 104 deletions(-) diff --git a/tests/test_models.py b/tests/test_models.py index e68e6599..b7831ef0 100644 --- a/tests/test_models.py +++ b/tests/test_models.py @@ -110,8 +110,10 @@ def test_model_forward_torchscript(model_name, batch_size): EXCLUDE_FEAT_FILTERS = [ 'hrnet*', '*pruned*', # hopefully fix at some point - 'legacy*', # not going to bother ] +if 'GITHUB_ACTIONS' in os.environ and 'Linux' in platform.system(): + # GitHub Linux runner is slower and hits memory limits sooner than MacOS, exclude bigger models + EXCLUDE_FEAT_FILTERS += ['*resnext101_32x32d'] @pytest.mark.timeout(120) @pytest.mark.parametrize('model_name', list_models(exclude_filters=EXCLUDE_FILTERS + EXCLUDE_FEAT_FILTERS)) diff --git a/timm/models/senet.py b/timm/models/senet.py index e4fca920..2156e4cd 100644 --- a/timm/models/senet.py +++ b/timm/models/senet.py @@ -18,7 +18,7 @@ import torch.nn as nn import torch.nn.functional as F from timm.data import IMAGENET_DEFAULT_MEAN, IMAGENET_DEFAULT_STD -from .helpers import load_pretrained +from .helpers import build_model_with_cfg from .layers import SelectAdaptivePool2d from .registry import register_model @@ -229,8 +229,8 @@ class SEResNetBlock(nn.Module): class SENet(nn.Module): def __init__(self, block, layers, groups, reduction, drop_rate=0.2, - in_chans=3, inplanes=128, input_3x3=True, downsample_kernel_size=3, - downsample_padding=1, num_classes=1000, global_pool='avg'): + in_chans=3, inplanes=64, input_3x3=False, downsample_kernel_size=1, + downsample_padding=0, num_classes=1000, global_pool='avg'): """ Parameters ---------- @@ -297,10 +297,10 @@ class SENet(nn.Module): ('bn1', nn.BatchNorm2d(inplanes)), ('relu1', nn.ReLU(inplace=True)), ] - # To preserve compatibility with Caffe weights `ceil_mode=True` - # is used instead of `padding=1`. - layer0_modules.append(('pool', nn.MaxPool2d(3, stride=2, ceil_mode=True))) self.layer0 = nn.Sequential(OrderedDict(layer0_modules)) + # To preserve compatibility with Caffe weights `ceil_mode=True` is used instead of `padding=1`. + self.pool0 = nn.MaxPool2d(3, stride=2, ceil_mode=True) + self.feature_info = [dict(num_chs=inplanes, reduction=2, module='layer0')] self.layer1 = self._make_layer( block, planes=64, @@ -310,6 +310,7 @@ class SENet(nn.Module): downsample_kernel_size=1, downsample_padding=0 ) + self.feature_info += [dict(num_chs=64 * block.expansion, reduction=4, module='layer1')] self.layer2 = self._make_layer( block, planes=128, @@ -320,6 +321,7 @@ class SENet(nn.Module): downsample_kernel_size=downsample_kernel_size, downsample_padding=downsample_padding ) + self.feature_info += [dict(num_chs=128 * block.expansion, reduction=8, module='layer2')] self.layer3 = self._make_layer( block, planes=256, @@ -330,6 +332,7 @@ class SENet(nn.Module): downsample_kernel_size=downsample_kernel_size, downsample_padding=downsample_padding ) + self.feature_info += [dict(num_chs=256 * block.expansion, reduction=16, module='layer3')] self.layer4 = self._make_layer( block, planes=512, @@ -340,8 +343,9 @@ class SENet(nn.Module): downsample_kernel_size=downsample_kernel_size, downsample_padding=downsample_padding ) - self.avg_pool = SelectAdaptivePool2d(pool_type=global_pool) + self.feature_info += [dict(num_chs=512 * block.expansion, reduction=32, module='layer4')] self.num_features = 512 * block.expansion + self.global_pool = SelectAdaptivePool2d(pool_type=global_pool) self.last_linear = nn.Linear(self.num_features, num_classes) for m in self.modules(): @@ -352,14 +356,13 @@ class SENet(nn.Module): downsample = None if stride != 1 or self.inplanes != planes * block.expansion: downsample = nn.Sequential( - nn.Conv2d(self.inplanes, planes * block.expansion, - kernel_size=downsample_kernel_size, stride=stride, - padding=downsample_padding, bias=False), + nn.Conv2d( + self.inplanes, planes * block.expansion, kernel_size=downsample_kernel_size, + stride=stride, padding=downsample_padding, bias=False), nn.BatchNorm2d(planes * block.expansion), ) - layers = [block( - self.inplanes, planes, groups, reduction, stride, downsample)] + layers = [block(self.inplanes, planes, groups, reduction, stride, downsample)] self.inplanes = planes * block.expansion for i in range(1, blocks): layers.append(block(self.inplanes, planes, groups, reduction)) @@ -371,15 +374,16 @@ class SENet(nn.Module): def reset_classifier(self, num_classes, global_pool='avg'): self.num_classes = num_classes - self.avg_pool = SelectAdaptivePool2d(pool_type=global_pool) + self.global_pool = SelectAdaptivePool2d(pool_type=global_pool) if num_classes: - num_features = self.num_features * self.avg_pool.feat_mult() + num_features = self.num_features * self.global_pool.feat_mult() self.last_linear = nn.Linear(num_features, num_classes) else: self.last_linear = nn.Identity() def forward_features(self, x): x = self.layer0(x) + x = self.pool0(x) x = self.layer1(x) x = self.layer2(x) x = self.layer3(x) @@ -387,7 +391,7 @@ class SENet(nn.Module): return x def logits(self, x): - x = self.avg_pool(x).flatten(1) + x = self.global_pool(x).flatten(1) if self.drop_rate > 0.: x = F.dropout(x, p=self.drop_rate, training=self.training) x = self.last_linear(x) @@ -399,116 +403,70 @@ class SENet(nn.Module): return x +def _create_senet(variant, pretrained=False, **kwargs): + return build_model_with_cfg( + SENet, variant, default_cfg=default_cfgs[variant], pretrained=pretrained, **kwargs) + + @register_model -def legacy_seresnet18(pretrained=False, num_classes=1000, in_chans=3, **kwargs): - default_cfg = default_cfgs['seresnet18'] - model = SENet(SEResNetBlock, [2, 2, 2, 2], groups=1, reduction=16, - inplanes=64, input_3x3=False, - downsample_kernel_size=1, downsample_padding=0, - num_classes=num_classes, in_chans=in_chans, **kwargs) - model.default_cfg = default_cfg - if pretrained: - load_pretrained(model, default_cfg, num_classes, in_chans) - return model +def legacy_seresnet18(pretrained=False, **kwargs): + model_args = dict( + block=SEResNetBlock, layers=[2, 2, 2, 2], groups=1, reduction=16, **kwargs) + return _create_senet('seresnet18', pretrained, **model_args) @register_model -def legacy_seresnet34(pretrained=False, num_classes=1000, in_chans=3, **kwargs): - default_cfg = default_cfgs['seresnet34'] - model = SENet(SEResNetBlock, [3, 4, 6, 3], groups=1, reduction=16, - inplanes=64, input_3x3=False, - downsample_kernel_size=1, downsample_padding=0, - num_classes=num_classes, in_chans=in_chans, **kwargs) - model.default_cfg = default_cfg - if pretrained: - load_pretrained(model, default_cfg, num_classes, in_chans) - return model +def legacy_seresnet34(pretrained=False, **kwargs): + model_args = dict( + block=SEResNetBlock, layers=[3, 4, 6, 3], groups=1, reduction=16, **kwargs) + return _create_senet('seresnet34', pretrained, **model_args) @register_model def legacy_seresnet50(pretrained=False, num_classes=1000, in_chans=3, **kwargs): - default_cfg = default_cfgs['seresnet50'] - model = SENet(SEResNetBottleneck, [3, 4, 6, 3], groups=1, reduction=16, - inplanes=64, input_3x3=False, - downsample_kernel_size=1, downsample_padding=0, - num_classes=num_classes, in_chans=in_chans, **kwargs) - model.default_cfg = default_cfg - if pretrained: - load_pretrained(model, default_cfg, num_classes, in_chans) - return model + model_args = dict( + block=SEResNetBottleneck, layers=[3, 4, 6, 3], groups=1, reduction=16, **kwargs) + return _create_senet('seresnet50', pretrained, **model_args) @register_model def legacy_seresnet101(pretrained=False, num_classes=1000, in_chans=3, **kwargs): - default_cfg = default_cfgs['seresnet101'] - model = SENet(SEResNetBottleneck, [3, 4, 23, 3], groups=1, reduction=16, - inplanes=64, input_3x3=False, - downsample_kernel_size=1, downsample_padding=0, - num_classes=num_classes, in_chans=in_chans, **kwargs) - model.default_cfg = default_cfg - if pretrained: - load_pretrained(model, default_cfg, num_classes, in_chans) - return model + model_args = dict( + block=SEResNetBottleneck, layers=[3, 4, 23, 3], groups=1, reduction=16, **kwargs) + return _create_senet('seresnet101', pretrained, **model_args) @register_model -def legacy_seresnet152(pretrained=False, num_classes=1000, in_chans=3, **kwargs): - default_cfg = default_cfgs['seresnet152'] - model = SENet(SEResNetBottleneck, [3, 8, 36, 3], groups=1, reduction=16, - inplanes=64, input_3x3=False, - downsample_kernel_size=1, downsample_padding=0, - num_classes=num_classes, in_chans=in_chans, **kwargs) - model.default_cfg = default_cfg - if pretrained: - load_pretrained(model, default_cfg, num_classes, in_chans) - return model +def legacy_seresnet152(pretrained=False, **kwargs): + model_args = dict( + block=SEResNetBottleneck, layers=[3, 8, 36, 3], groups=1, reduction=16, **kwargs) + return _create_senet('seresnet152', pretrained, **model_args) @register_model -def legacy_senet154(pretrained=False, num_classes=1000, in_chans=3, **kwargs): - default_cfg = default_cfgs['senet154'] - model = SENet(SEBottleneck, [3, 8, 36, 3], groups=64, reduction=16, - num_classes=num_classes, in_chans=in_chans, **kwargs) - model.default_cfg = default_cfg - if pretrained: - load_pretrained(model, default_cfg, num_classes, in_chans) - return model +def legacy_senet154(pretrained=False, **kwargs): + model_args = dict( + block=SEBottleneck, layers=[3, 8, 36, 3], groups=64, reduction=16, + downsample_kernel_size=3, downsample_padding=1, inplanes=128, input_3x3=True, **kwargs) + return _create_senet('senet154', pretrained, **model_args) @register_model -def legacy_seresnext26_32x4d(pretrained=False, num_classes=1000, in_chans=3, **kwargs): - default_cfg = default_cfgs['seresnext26_32x4d'] - model = SENet(SEResNeXtBottleneck, [2, 2, 2, 2], groups=32, reduction=16, - inplanes=64, input_3x3=False, - downsample_kernel_size=1, downsample_padding=0, - num_classes=num_classes, in_chans=in_chans, **kwargs) - model.default_cfg = default_cfg - if pretrained: - load_pretrained(model, default_cfg, num_classes, in_chans) - return model +def legacy_seresnext26_32x4d(pretrained=False, **kwargs): + model_args = dict( + block=SEResNeXtBottleneck, layers=[2, 2, 2, 2], groups=32, reduction=16, **kwargs) + return _create_senet('seresnext26_32x4d', pretrained, **model_args) @register_model -def legacy_seresnext50_32x4d(pretrained=False, num_classes=1000, in_chans=3, **kwargs): - default_cfg = default_cfgs['seresnext50_32x4d'] - model = SENet(SEResNeXtBottleneck, [3, 4, 6, 3], groups=32, reduction=16, - inplanes=64, input_3x3=False, - downsample_kernel_size=1, downsample_padding=0, - num_classes=num_classes, in_chans=in_chans, **kwargs) - model.default_cfg = default_cfg - if pretrained: - load_pretrained(model, default_cfg, num_classes, in_chans) - return model +def legacy_seresnext50_32x4d(pretrained=False, **kwargs): + model_args = dict( + block=SEResNeXtBottleneck, layers=[3, 4, 6, 3], groups=32, reduction=16, **kwargs) + return _create_senet('seresnext50_32x4d', pretrained, **model_args) @register_model -def legacy_seresnext101_32x4d(pretrained=False, num_classes=1000, in_chans=3, **kwargs): - default_cfg = default_cfgs['seresnext101_32x4d'] - model = SENet(SEResNeXtBottleneck, [3, 4, 23, 3], groups=32, reduction=16, - inplanes=64, input_3x3=False, - downsample_kernel_size=1, downsample_padding=0, - num_classes=num_classes, in_chans=in_chans, **kwargs) - model.default_cfg = default_cfg - if pretrained: - load_pretrained(model, default_cfg, num_classes, in_chans) - return model +def legacy_seresnext101_32x4d(pretrained=False, **kwargs): + model_args = dict( + block=SEResNeXtBottleneck, layers=[3, 4, 23, 3], groups=32, reduction=16, **kwargs) + return _create_senet('seresnext101_32x4d', pretrained, **model_args) diff --git a/timm/models/xception_aligned.py b/timm/models/xception_aligned.py index 75ba7a27..8303af27 100644 --- a/timm/models/xception_aligned.py +++ b/timm/models/xception_aligned.py @@ -1,6 +1,7 @@ -"""Pytorch impl of Aligned Xception +"""Pytorch impl of Aligned Xception 41, 65, 71 -This is a correct impl of Aligned Xception (Deeplab) models compatible with TF definition. +This is a correct, from scratch impl of Aligned Xception (Deeplab) models compatible with TF weights at +https://github.com/tensorflow/models/blob/master/research/deeplab/g3doc/model_zoo.md Hacked together by Ross Wightman """