monk-cls-test1 0.0.5

Monk Classification's Gluoncv backend

List of all available functions and associated tutorials

• Available Backend frameworks
• Available Transfer Learning Models
• Available Layers
• Available Activation Functions
• Available Optimizers
• Available Loss functions
• Available network blocks

Available backend frameworks

a) Mxnet Gluon - version 1.5.1
b) Pytorch     - version 1.2.0
c) Keras       - version 2.2.5 (tf - 1.12.0)

Available Transfer Learning Models

monk name	Original Name in Keras	Original Name in Pytorch	Original Name in MXNet
alexnet	-	alexnet	AlexNet
darknet	-	-	Darnet53
densenet121	DenseNet121	densenet121	DenseNet121
densenet161	-	densenet161	DenseNet161
densenet169	DenseNet169	densenet169	DenseNet169
densenet201	DenseNet201	densenet201	DenseNet201
googlenet	-	googlenet	-
inception_v3	InceptionV3	inception_v3	InceptionV3
inception_resnet_v2	InceptionResNetV2	-	-
mnasnet0_5	-	mnasnet0_5	-
mnasnet0_75	-	mnasnet0_75	-
mnasnet1_0	-	mnasnet1_0	-
mnasnet1_3	-	mnasnet1_3	-
nasnet_mobile	NASNetMobile	-	-
nasnet_large	NASNetLarge	-	-
mobilenet	MobileNet	-	MobileNet1.0
mobilenet1.0_int8	-	-	MobileNet1.0_int8
mobilenet0.75	-	-	MobileNet0.75
mobilenet0.5	-	-	MobileNet0.5
mobilenet0.25	-	-	MobileNet0.25
mobilenetv2	MobileNetV2	mobilenet_v2	MobileNetV2_1.0
mobilenetv2_0.75	-	-	MobileNetV2_0.75
mobilenetv2_0.5	-	-	MobileNetV2_0.5
mobilenetv2_0.25	-	-	MobileNetV2_0.25
mobilenetv3_large	-	-	MobileNetV3_Large
mobilenetv3_smalle	-	-	MobileNetV3_Small
resnet18_v1	-	resnet18	ResNet18_v1
resnet34_v1	-	resnet34	ResNet34_v1
resnet50_v1	ResNet50	resnet50	ResNet50_v1
resnet101_v1	ResNet101	resnet101	ResNet101_v1
resnet152_v1	ResNet152	resnet152	ResNet152_v1
resnet18_v2			ResNet18_v2
resnet34_v2			ResNet34_v2
resnet50_v2	ResNet50V2	-	ResNet50_v2
resnet101_v2	ResNet101V2	-	ResNet101_v2
resnet152_v2	ResNet152V2	-	ResNet152_v2
resnext50_32x4d	-	resnext50_32x4d	ResNext50_32x4d
resnext101_32x8d	-	resnext101_32x8d	ResNext101_32x4d
resnext101_64x4d	-	-	ResNext101_64x4d
se_resnext50_32x4d	-	-	SE_ResNext50_32x4d
se_resnext101_32x4d	-	-	SE_ResNext101_32x4d
se_resnext101_64x4d	-	-	SE_ResNext101_64x4d
shufflenet_v2_x0_5	-	shufflenet_v2_x0_5	-
shufflenet_v2_x1_0	-	shufflenet_v2_x1_0	-
shufflenet_v2_x1_5	-	shufflenet_v2_x1_5	-
shufflenet_v2_x2_0	-	shufflenet_v2_x2_0	-
squeezenet1_0	-	squeezenet1_0	SqueezeNet1.0
squeezenet1_1	-	squeezenet1_1	SqueezeNet1.1
senet_154	-	-	SENet_154
vgg11	-	vgg11	VGG11
vgg11_bn	-	vgg11_bn	VGG11_bn
vgg13	-	vgg13	VGG13
vgg13_bn	-	vgg13_bn	VGG13_bn
vgg16	VGG16	vgg16	VGG16
vgg16_bn	-	vgg16_bn	VGG16_bn
vgg19	VGG19	vgg19	VGG19
vgg19_bn	-	vgg19_bn	VGG19_bn
wide_resnet50_2	-	wide_resnet50_2	-
wide_resnet101_2	-	wide_resnet101_2	-
xception	Xception	-	Xception

Available Custom Network Layers

Name in Monk	Name in Keras backend	Name in pytorch backend	Name in mxnet backed
fully_connected	Dense	Linear	Dense
Dropout	Dropout	Dropout	Dropout
Flatten	Flatten	Flatten	Flatten
convolution1d	Conv1D	Conv1d	Conv1D
convolution	Conv2D	Conv2d	Conv2D
convolution3d	Conv3D	Conv3d	Conv3D
transposed_convolution1d	-	ConvTranspose1d	Conv1DTranspose
transposed_convolution	Conv2DTranspose	ConvTranspose2d	Conv2DTranspose
transposed_convolution3d	Conv3DTranspose	ConvTranspose3d	Conv3DTranspose
max_pooling1d	MaxPooling1D	MaxPool1d	MaxPool1D
max_pooling	MaxPooling2D	MaxPool2d	MaxPool2D
max_pooling3d	MaxPooling3D	MaxPool3d	MaxPool3D
average_pooling1d	AveragePooling1D	AvgPool1d	AvgPool1D
average_pooling	AveragePooling2D	AvgPool2d	AvgPool2D
average_pooling3d	AveragePooling3D	AvgPool3d	AvgPool3D
global_max_pooling1d	GlobalMaxPooling1D	AdaptiveMaxPool1d (With size = 1)	GlobalMaxPool1D
global_max_pooling	GlobalMaxPooling2D	AdaptiveMaxPool2d (With size = 1)	GlobalMaxPool2D
global_max_pooling3d	GlobalMaxPooling3D	AdaptiveMaxPool3d (With size = 1)	GlobalMaxPool3D
global_average_pooling1d	GlobalAveragePooling1D	AdaptiveAvgPool1d (With size = 1)	GlobalAvgPool1D
global_average_pooling	GlobalAveragePooling2D	AdaptiveAvgPool2d (With size = 1)	GlobalAvgPool2D
global_average_pooling3d	GlobalAveragePooling3D	AdaptiveAvgPool3d (With size = 1)	GlobalAvgPool3D
add	Add	Add	Add
concatenate	Concatenate	Concatenate	Concatenate
batchnorm	BatchNormalization	BatchNorm1d	BatchNorm
batchnorm	-	BatchNorm2d	-
batchnorm	-	BatchNorm3d	-
instancenorm	-	InstanceNorm1d	InstanceNorm
instancenorm	-	InstanceNorm2d	-
instancenorm	-	InstanceNorm3d	-
layernorm	-	LayerNorm	LayerNorm
identity	activation.linear	Identity	Identity

Available Custom Network Activation Functions

Name in Monk	Original name in Keras backend	Original name in pytorch backend	Original name in mxnet backend
relu	relu	ReLU	Activation('relu')
sigmoid	sigmoid	Sigmoid	Activation('sigmoid')
Tanh Shrink	tanh	TanH	Activation('tanh')
softplus	softplus	Softplus	Activation('softrelu')
softsign	softsign	Softsign	Activation('softsign')
elu	elu	ELU	ELU
gelu	-	-	GELU
prelu	PReLU	PReLU	PReLU
selu	selu	SELU	SELU
swish	-	-	Swish
leakyrelu	LeakyReLU	LeakyReLU	LeakyReLU
hardshrink	-	HardShrink	-
hardtanh	-	HardTanh	-
logsigmoid	-	LogSigmoid	-
relu6	-	ReLU6	-
rrelu	-	RReLU	-
celu	-	CELU	-
softshrink	-	Softshrink	-
tanhshrink	-	Tanhshrink	-
threshold	-	Threshold	-
softmin	-	Softmin	-
softmax	-	Softmax	-
logsoftmax	-	LogSoftmax	-
hardsigmoid	hard_sigmoid	-	-
thresholded_relu	ThresholdedReLU	-	-

Available Optimizers

Name in Monk	Original Name in Keras backend	Original Name in pytorch backend	Original Name in mxnet backend
optimizer_adadelta	Adadelta	Adadelta	AdaDelta
optimizer_adagrad	Adagrad	Adagrad	AdaGrad
optimizer_adam	Adam	Adam	Adam
optimizer_adamax	Adamax	Adamax	Adamax
optimizer_nesterov_sgd	SGD (With nesterov)	SGD (With nesterov)	NAG
optimizer_nesterov_adam	Nadam	-	Nadam
optimizer_rmsprop	RMSprop	RMSprop	RMSProp
optimizer_momentum_rmsprop	-	RMSprop (With momentum)	RMSprop (With momentum)
optimizer_sgd	SGD	SGD	SGD
optimizer_signum	-	-	Signum
optimizer_adamw	-	AdamW	-

Available Loss functions

Name in Monk	Original Name in keras backend	Original Name in pytorch backend	Original Name in mxnet backend
loss_l2	mean_squared_error	MSELoss	L2Loss
loss_l1	mean_absolute_error	L1Loss	L1Loss
loss_squared_hinge	squared_hinge	SoftMarginLoss (not exactly)	SquaredHingeLoss
loss_hinge	hinge	HingeEmbeddingLoss	HingeLoss
loss_huber	huber_loss	SmoothL1Loss	HuberLoss
loss_softmax_crossentropy	-	CrossEntropyLoss	SoftmaxCrossEntropyLoss
loss_crossentropy	categorical_crossentropy	CrossEntropyLoss	SoftmaxCrossEntropyLoss
loss_multimargin	categorical_hinge	MultiMarginLoss	-
loss_multilabel_margin	-	MultiLabelMarginLoss	-
loss_binary_crossentropy	binary_crossentropy	BCELoss	-
loss_sigmoid_binary_crossentropy	-	BCEWithLogitsLoss	SigmoidBinaryCrossEntropyLoss
loss_kldiv	kullback_leibler_divergence	KLDivLoss	KLDivLoss
loss_poison_nll	-	PoissonNLLLoss	PoissonNLLLoss

Available network blocks

Block	Name in Monk
Resnet V1 Block With Downsampling	resnet_v1_block
Resnet V1 Block Without Downsampling	resnet_v1_block
Resnet V2 Block With Downsampling	resnet_v2_block
Resnet V2 Block Without Downsampling	resnet_v2_block
Resnet V1 Bottleneck Block With Downsampling	resnet_v1_bottleneck_block
Resnet V1 Bottleneck Block Without Downsampling	resnet_v1_bottleneck_block
Resnet V2 Bottleneck Block With Downsampling	resnet_v2_bottleneck_block
Resnet V2 Bottleneck Block Without Downsampling	resnet_v2_bottleneck_block
Resnext Block With Downsampling	resnext_block
Resnext Block Without Downsampling	resnext_block
Mobilenet V2 Linear BottleNeck Block	mobilenet_v2_linear_block
Mobilenet V2 Inverted Linear BottleNeck Block	mobilenet_v2_inverted_linear_block
Squeezenet Fire Block	squeezenet_fire_block
Densenet Dense Block	densenet_dense_block
Inception A Block	inception_a_block
Inception B Block	inception_b_block
Inception C Block	inception_c_block
Inception D Block	inception_d_block
Inception E Block	inception_e_block