force mxnet>=1.3.0, replace load_params with load_parameters (#398)

* force mxnet>=1.3.0, replace load_params with load_parameters

* fix pylint

docs/tutorials/classification/dive_deep_cifar10.py

@@ -317,7 +317,7 @@ def test(ctx, val_data):
 ################################################################
 # Next time if you need to use it, just run
 
-net.load_params('dive_deep_cifar10_resnet20_v2.params', ctx=ctx)
+net.load_parameters('dive_deep_cifar10_resnet20_v2.params', ctx=ctx)
 
 ################################################################
 # Next Step

gluoncv/__init__.py

@@ -7,18 +7,17 @@
 mx_version = '1.3.0'
 try:
     import mxnet as mx
-    import warnings
     from distutils.version import LooseVersion
     if LooseVersion(mx.__version__) < LooseVersion(mx_version):
         msg = (
             "Legacy mxnet=={} detected, some new modules may not work properly. "
-            "mxnet>={} is recommended. You can use pip to upgrade mxnet "
-            "`pip install mxnet --pre --upgrade`").format(mx.__version__, mx_version)
-        warnings.warn(msg)
+            "mxnet>={} is required. You can use pip to upgrade mxnet "
+            "`pip install mxnet/mxnet-cu90 --pre --upgrade`").format(mx.__version__, mx_version)
+        raise ImportError(msg)
 except ImportError:
     raise ImportError(
         "Unable to import dependency mxnet. "
-        "A quick tip is to install via `pip install mxnet --pre`. "
+        "A quick tip is to install via `pip install mxnet/mxnet-cu90 --pre`. "
         "please refer to https://gluon-cv.mxnet.io/#installation for details.")
 
 __version__ = '0.3.0'

gluoncv/model_zoo/cifarresnet.py

@@ -269,8 +269,8 @@ def get_cifar_resnet(version, num_layers, pretrained=False, ctx=cpu(),
     net = resnet_class(block_class, layers, channels, **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        net.load_params(get_model_file('cifar_resnet%d_v%d'%(num_layers, version),
-                                       tag=pretrained, root=root), ctx=ctx)
+        net.load_parameters(get_model_file('cifar_resnet%d_v%d'%(num_layers, version),
+                                           tag=pretrained, root=root), ctx=ctx)
     return net
 
 def cifar_resnet20_v1(**kwargs):

gluoncv/model_zoo/cifarresnext.py

@@ -163,9 +163,9 @@ def get_cifar_resnext(num_layers, cardinality=16, bottleneck_width=64,
     net = CIFARResNext(layers, cardinality, bottleneck_width, **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        net.load_params(get_model_file('cifar_resnext%d_%dx%dd'%(num_layers, cardinality,
-                                                                 bottleneck_width),
-                                       tag=pretrained, root=root), ctx=ctx)
+        net.load_parameters(get_model_file('cifar_resnext%d_%dx%dd'%(num_layers, cardinality,
+                                                                     bottleneck_width),
+                                           tag=pretrained, root=root), ctx=ctx)
     return net
 
 def cifar_resnext29_32x4d(**kwargs):

gluoncv/model_zoo/cifarwideresnet.py

@@ -166,8 +166,8 @@ def get_cifar_wide_resnet(num_layers, width_factor=1, drop_rate=0.0,
     net = CIFARWideResNet(CIFARBasicBlockV2, layers, channels, drop_rate, **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        net.load_params(get_model_file('cifar_wideresnet%d_%d'%(num_layers, width_factor),
-                                       tag=pretrained, root=root), ctx=ctx)
+        net.load_parameters(get_model_file('cifar_wideresnet%d_%d'%(num_layers, width_factor),
+                                           tag=pretrained, root=root), ctx=ctx)
     return net
 
 def cifar_wideresnet16_10(**kwargs):

gluoncv/model_zoo/nasnet.py

@@ -620,8 +620,8 @@ def get_nasnet(repeat=6, penultimate_filters=4032,
     net = NASNetALarge(repeat=repeat, penultimate_filters=penultimate_filters, **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        net.load_params(get_model_file('nasnet_%d_%d'%(repeat, penultimate_filters),
-                                       tag=pretrained, root=root), ctx=ctx)
+        net.load_parameters(get_model_file('nasnet_%d_%d'%(repeat, penultimate_filters),
+                                           tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         net.synset = attrib.synset

gluoncv/model_zoo/resnetv1b.py

@@ -279,8 +279,8 @@ def resnet18_v1b(pretrained=False, root='~/.mxnet/models', ctx=cpu(0), **kwargs)
     model = ResNetV1b(BasicBlockV1b, [2, 2, 2, 2], **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        model.load_params(get_model_file('resnet%d_v%db'%(18, 1),
-                                         tag=pretrained, root=root), ctx=ctx)
+        model.load_parameters(get_model_file('resnet%d_v%db'%(18, 1),
+                                             tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         model.synset = attrib.synset
@@ -314,8 +314,8 @@ def resnet34_v1b(pretrained=False, root='~/.mxnet/models', ctx=cpu(0), **kwargs)
     model = ResNetV1b(BasicBlockV1b, [3, 4, 6, 3], **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        model.load_params(get_model_file('resnet%d_v%db'%(34, 1),
-                                         tag=pretrained, root=root), ctx=ctx)
+        model.load_parameters(get_model_file('resnet%d_v%db'%(34, 1),
+                                             tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         model.synset = attrib.synset
@@ -349,8 +349,8 @@ def resnet50_v1b(pretrained=False, root='~/.mxnet/models', ctx=cpu(0), **kwargs)
     model = ResNetV1b(BottleneckV1b, [3, 4, 6, 3], **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        model.load_params(get_model_file('resnet%d_v%db'%(50, 1),
-                                         tag=pretrained, root=root), ctx=ctx)
+        model.load_parameters(get_model_file('resnet%d_v%db'%(50, 1),
+                                             tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         model.synset = attrib.synset
@@ -384,8 +384,8 @@ def resnet101_v1b(pretrained=False, root='~/.mxnet/models', ctx=cpu(0), **kwargs
     model = ResNetV1b(BottleneckV1b, [3, 4, 23, 3], **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        model.load_params(get_model_file('resnet%d_v%db'%(101, 1),
-                                         tag=pretrained, root=root), ctx=ctx)
+        model.load_parameters(get_model_file('resnet%d_v%db'%(101, 1),
+                                             tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         model.synset = attrib.synset
@@ -419,8 +419,8 @@ def resnet152_v1b(pretrained=False, root='~/.mxnet/models', ctx=cpu(0), **kwargs
     model = ResNetV1b(BottleneckV1b, [3, 8, 36, 3], **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        model.load_params(get_model_file('resnet%d_v%db'%(152, 1),
-                                         tag=pretrained, root=root), ctx=ctx)
+        model.load_parameters(get_model_file('resnet%d_v%db'%(152, 1),
+                                             tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         model.synset = attrib.synset
@@ -448,8 +448,8 @@ def resnet50_v1c(pretrained=False, root='~/.mxnet/models', ctx=cpu(0), **kwargs)
     model = ResNetV1b(BottleneckV1b, [3, 4, 6, 3], deep_stem=True, **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        model.load_params(get_model_file('resnet%d_v%dc'%(50, 1),
-                                         tag=pretrained, root=root), ctx=ctx)
+        model.load_parameters(get_model_file('resnet%d_v%dc'%(50, 1),
+                                             tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         model.synset = attrib.synset
@@ -478,8 +478,8 @@ def resnet101_v1c(pretrained=False, root='~/.mxnet/models', ctx=cpu(0), **kwargs
     model = ResNetV1b(BottleneckV1b, [3, 4, 23, 3], deep_stem=True, **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        model.load_params(get_model_file('resnet%d_v%dc'%(101, 1),
-                                         tag=pretrained, root=root), ctx=ctx)
+        model.load_parameters(get_model_file('resnet%d_v%dc'%(101, 1),
+                                             tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         model.synset = attrib.synset
@@ -507,8 +507,8 @@ def resnet152_v1c(pretrained=False, root='~/.mxnet/models', ctx=cpu(0), **kwargs
     model = ResNetV1b(BottleneckV1b, [3, 8, 36, 3], deep_stem=True, **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        model.load_params(get_model_file('resnet%d_v%dc'%(152, 1),
-                                         tag=pretrained, root=root), ctx=ctx)
+        model.load_parameters(get_model_file('resnet%d_v%dc'%(152, 1),
+                                             tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         model.synset = attrib.synset
@@ -536,8 +536,8 @@ def resnet50_v1d(pretrained=False, root='~/.mxnet/models', ctx=cpu(0), **kwargs)
     model = ResNetV1b(BottleneckV1b, [3, 4, 6, 3], deep_stem=True, avg_down=True, **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        model.load_params(get_model_file('resnet%d_v%dd'%(50, 1),
-                                         tag=pretrained, root=root), ctx=ctx)
+        model.load_parameters(get_model_file('resnet%d_v%dd'%(50, 1),
+                                             tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         model.synset = attrib.synset
@@ -565,8 +565,8 @@ def resnet101_v1d(pretrained=False, root='~/.mxnet/models', ctx=cpu(0), **kwargs
     model = ResNetV1b(BottleneckV1b, [3, 4, 23, 3], deep_stem=True, avg_down=True, **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        model.load_params(get_model_file('resnet%d_v%dd'%(101, 1),
-                                         tag=pretrained, root=root), ctx=ctx)
+        model.load_parameters(get_model_file('resnet%d_v%dd'%(101, 1),
+                                             tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         model.synset = attrib.synset
@@ -594,8 +594,8 @@ def resnet152_v1d(pretrained=False, root='~/.mxnet/models', ctx=cpu(0), **kwargs
     model = ResNetV1b(BottleneckV1b, [3, 8, 36, 3], deep_stem=True, avg_down=True, **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        model.load_params(get_model_file('resnet%d_v%dd'%(152, 1),
-                                         tag=pretrained, root=root), ctx=ctx)
+        model.load_parameters(get_model_file('resnet%d_v%dd'%(152, 1),
+                                             tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         model.synset = attrib.synset
@@ -624,8 +624,8 @@ def resnet50_v1e(pretrained=False, root='~/.mxnet/models', ctx=cpu(0), **kwargs)
                       deep_stem=True, avg_down=True, stem_width=64, **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        model.load_params(get_model_file('resnet%d_v%dd'%(50, 1),
-                                         tag=pretrained, root=root), ctx=ctx)
+        model.load_parameters(get_model_file('resnet%d_v%dd'%(50, 1),
+                                             tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         model.synset = attrib.synset
@@ -654,8 +654,8 @@ def resnet101_v1e(pretrained=False, root='~/.mxnet/models', ctx=cpu(0), **kwargs
                       deep_stem=True, avg_down=True, stem_width=64, **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        model.load_params(get_model_file('resnet%d_v%dd'%(101, 1),
-                                         tag=pretrained, root=root), ctx=ctx)
+        model.load_parameters(get_model_file('resnet%d_v%dd'%(101, 1),
+                                             tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         model.synset = attrib.synset
@@ -684,8 +684,8 @@ def resnet152_v1e(pretrained=False, root='~/.mxnet/models', ctx=cpu(0), **kwargs
                       deep_stem=True, avg_down=True, stem_width=64, **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        model.load_params(get_model_file('resnet%d_v%dd'%(152, 1),
-                                         tag=pretrained, root=root), ctx=ctx)
+        model.load_parameters(get_model_file('resnet%d_v%dd'%(152, 1),
+                                             tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         model.synset = attrib.synset
@@ -714,8 +714,8 @@ def resnet50_v1s(pretrained=False, root='~/.mxnet/models', ctx=cpu(0), **kwargs)
     model = ResNetV1b(BottleneckV1b, [3, 4, 6, 3], deep_stem=True, stem_width=64, **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        model.load_params(get_model_file('resnet%d_v%ds'%(50, 1),
-                                         tag=pretrained, root=root), ctx=ctx)
+        model.load_parameters(get_model_file('resnet%d_v%ds'%(50, 1),
+                                             tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         model.synset = attrib.synset
@@ -744,8 +744,8 @@ def resnet101_v1s(pretrained=False, root='~/.mxnet/models', ctx=cpu(0), **kwargs
     model = ResNetV1b(BottleneckV1b, [3, 4, 23, 3], deep_stem=True, stem_width=64, **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        model.load_params(get_model_file('resnet%d_v%ds'%(101, 1),
-                                         tag=pretrained, root=root), ctx=ctx)
+        model.load_parameters(get_model_file('resnet%d_v%ds'%(101, 1),
+                                             tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         model.synset = attrib.synset
@@ -774,8 +774,8 @@ def resnet152_v1s(pretrained=False, root='~/.mxnet/models', ctx=cpu(0), **kwargs
     model = ResNetV1b(BottleneckV1b, [3, 8, 36, 3], deep_stem=True, stem_width=64, **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        model.load_params(get_model_file('resnet%d_v%ds'%(152, 1),
-                                         tag=pretrained, root=root), ctx=ctx)
+        model.load_parameters(get_model_file('resnet%d_v%ds'%(152, 1),
+                                             tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         model.synset = attrib.synset

gluoncv/model_zoo/resnext.py

@@ -203,13 +203,13 @@ def get_resnext(num_layers, cardinality=32, bottleneck_width=4, use_se=False,
     if pretrained:
         from .model_store import get_model_file
         if not use_se:
-            net.load_params(get_model_file('resnext%d_%dx%dd'%(num_layers, cardinality,
-                                                               bottleneck_width),
-                                           tag=pretrained, root=root), ctx=ctx)
+            net.load_parameters(get_model_file('resnext%d_%dx%dd'%(num_layers, cardinality,
+                                                                   bottleneck_width),
+                                               tag=pretrained, root=root), ctx=ctx)
         else:
-            net.load_params(get_model_file('se_resnext%d_%dx%dd'%(num_layers, cardinality,
-                                                                  bottleneck_width),
-                                           tag=pretrained, root=root), ctx=ctx)
+            net.load_parameters(get_model_file('se_resnext%d_%dx%dd'%(num_layers, cardinality,
+                                                                      bottleneck_width),
+                                               tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         net.synset = attrib.synset

gluoncv/model_zoo/se_resnet.py

@@ -432,8 +432,8 @@ def get_se_resnet(version, num_layers, pretrained=False, ctx=cpu(),
     net = resnet_class(block_class, layers, channels, **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        net.load_params(get_model_file('se_resnet%d_v%d'%(num_layers, version),
-                                       tag=pretrained, root=root), ctx=ctx)
+        net.load_parameters(get_model_file('se_resnet%d_v%d'%(num_layers, version),
+                                           tag=pretrained, root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         net.synset = attrib.synset

gluoncv/model_zoo/senet.py

@@ -195,8 +195,8 @@ def get_senet(num_layers, cardinality=64, bottleneck_width=4,
     net = SENet(layers, cardinality, bottleneck_width, **kwargs)
     if pretrained:
         from .model_store import get_model_file
-        net.load_params(get_model_file('senet_%d'%(num_layers+2),
-                                       root=root), ctx=ctx)
+        net.load_parameters(get_model_file('senet_%d'%(num_layers+2),
+                                           root=root), ctx=ctx)
         from ..data import ImageNet1kAttr
         attrib = ImageNet1kAttr()
         net.synset = attrib.synset

gluoncv/model_zoo/ssd/ssd.py

@@ -299,7 +299,7 @@ def get_ssd(name, base_size, features, filters, sizes, ratios, steps, classes,
     if pretrained:
         from ..model_store import get_model_file
         full_name = '_'.join(('ssd', str(base_size), name, dataset))
-        net.load_params(get_model_file(full_name, tag=pretrained, root=root), ctx=ctx)
+        net.load_parameters(get_model_file(full_name, tag=pretrained, root=root), ctx=ctx)
     return net
 
 def ssd_300_vgg16_atrous_voc(pretrained=False, pretrained_base=True, **kwargs):

gluoncv/model_zoo/ssd/vgg_atrous.py

@@ -196,8 +196,8 @@ def get_vgg_atrous_extractor(num_layers, im_size, pretrained=False, ctx=mx.cpu()
         from ..model_store import get_model_file
         batch_norm_suffix = '_bn' if kwargs.get('batch_norm') else ''
         net.initialize(ctx=ctx)
-        net.load_params(get_model_file('vgg%d_atrous%s'%(num_layers, batch_norm_suffix),
-                                       tag=pretrained, root=root), ctx=ctx, allow_missing=True)
+        net.load_parameters(get_model_file('vgg%d_atrous%s'%(num_layers, batch_norm_suffix),
+                                           tag=pretrained, root=root), ctx=ctx, allow_missing=True)
     return net
 
 def vgg16_atrous_300(**kwargs):

gluoncv/model_zoo/yolo/yolo3.py

@@ -475,7 +475,7 @@ def get_yolov3(name, stages, filters, anchors, strides, classes,
     if pretrained:
         from ..model_store import get_model_file
         full_name = '_'.join(('yolo3', name, dataset))
-        net.load_params(get_model_file(full_name, tag=pretrained, root=root), ctx=ctx)
+        net.load_parameters(get_model_file(full_name, tag=pretrained, root=root), ctx=ctx)
     return net
 
 def yolo3_darknet53_voc(pretrained_base=True, pretrained=False, num_sync_bn_devices=-1, **kwargs):

scripts/classification/cifar/demo_cifar10.py

@@ -31,7 +31,7 @@
 net = get_model(model_name, **kwargs)
 
 if not pretrained:
-    net.load_params(opt.saved_params, ctx = context)
+    net.load_parameters(opt.saved_params, ctx = context)
 
 # Load Images
 img = image.imread(opt.input_pic)

scripts/classification/cifar/train_cifar10.py

@@ -72,7 +72,7 @@
     kwargs = {'classes': classes}
 net = get_model(model_name, **kwargs)
 if opt.resume_from:
-    net.load_params(opt.resume_from, ctx = context)
+    net.load_parameters(opt.resume_from, ctx = context)
 optimizer = 'nag'
 
 save_period = opt.save_period