Gluon
search
Quick search
code
Show Source
D2L Book
GitHub
Table Of Contents
Guide
Getting started with NP on MXNet
Step 1: Manipulate data with NP on MXNet
Step 2: Create a neural network
Step 3: Automatic differentiation with autograd
Step 4: Train the neural network
Step 5: Predict with a pretrained model
Step 6: Use GPUs to increase efficiency
What is NP on MXNet
API
NP on MXNet reference
Array objects
The N-dimensional array (
ndarray
)
ndarray
ndarray.shape
ndarray.ndim
ndarray.size
ndarray.dtype
ndarray.T
ndarray.item
ndarray.copy
ndarray.tolist
ndarray.astype
ndarray.reshape
ndarray.transpose
ndarray.swapaxes
ndarray.flatten
ndarray.squeeze
ndarray.nonzero
ndarray.take
ndarray.repeat
ndarray.max
ndarray.argmax
ndarray.min
ndarray.argmin
ndarray.clip
ndarray.sum
ndarray.mean
ndarray.prod
ndarray.cumsum
ndarray.var
ndarray.std
ndarray.__lt__
ndarray.__le__
ndarray.__gt__
ndarray.__ge__
ndarray.__eq__
ndarray.__ne__
ndarray.__bool__
ndarray.__neg__
ndarray.__add__
ndarray.__sub__
ndarray.__mul__
ndarray.__truediv__
ndarray.__mod__
ndarray.__pow__
ndarray.__iadd__
ndarray.__isub__
ndarray.__imul__
ndarray.__itruediv__
ndarray.__imod__
ndarray.__reduce__
ndarray.__setstate__
ndarray.__len__
ndarray.__getitem__
ndarray.__setitem__
ndarray.__int__
ndarray.__float__
ndarray.__str__
ndarray.__repr__
Indexing
Routines
Array creation routines
eye
empty
full
identity
ones
ones_like
zeros
zeros_like
array
copy
arange
linspace
logspace
meshgrid
tril
Array manipulation routines
reshape
ravel
ndarray.flatten
swapaxes
ndarray.T
transpose
moveaxis
expand_dims
squeeze
broadcast_to
broadcast_arrays
concatenate
stack
dstack
vstack
split
hsplit
vsplit
tile
repeat
unique
reshape
flip
roll
rot90
Input and output
genfromtxt
Linear algebra (
numpy.linalg
)
dot
vdot
inner
outer
tensordot
einsum
svd
norm
trace
Mathematical functions
sin
cos
tan
arcsin
arccos
arctan
degrees
radians
hypot
arctan2
deg2rad
rad2deg
sinh
cosh
tanh
arcsinh
arccosh
arctanh
rint
fix
floor
ceil
trunc
around
sum
prod
cumsum
exp
expm1
log
log10
log2
log1p
ldexp
lcm
add
reciprocal
negative
divide
power
subtract
mod
multiply
true_divide
remainder
clip
sqrt
cbrt
square
absolute
sign
maximum
minimum
np.random
choice
shuffle
normal
uniform
rand
randint
Sorting, searching, and counting
argmax
argmin
Statistics
min
max
mean
std
var
histogram
NP Extensions
set_np
reset_np
cpu
cpu_pinned
gpu
gpu_memory_info
current_context
num_gpus
activation
batch_norm
convolution
dropout
embedding
fully_connected
layer_norm
pooling
rnn
leaky_relu
multibox_detection
multibox_prior
multibox_target
roi_pooling
sigmoid
smooth_l1
softmax
mxnet.npx.threading
topk
waitall
load
save
one_hot
pick
reshape_like
batch_flatten
batch_dot
gamma
sequence_mask
Gluon reference
Creating Nerual Networks
Block objects
Block
Block.forward
Block.__call__
Block.initialize
Block.save_parameters
Block.load_parameters
Block.collect_params
Block.cast
Block.apply
Block.summary
Block.name_scope
Block.register_child
Block.register_forward_hook
Block.register_forward_pre_hook
Block.name
Block.params
Block.prefix
Block.save_params
Block.load_params
HybridBlock
HybridBlock.hybridize
HybridBlock.hybrid_forward
SymbolBlock
Container objects
Sequential
HybridSequential
Concurrent
HybridConcurrent
Feedforward Layers
Dense
Activation
Dropout
Flatten
Lambda
HybridLambda
Conv1D
Conv2D
Conv3D
Conv1DTranspose
Conv2DTranspose
Conv3DTranspose
MaxPool1D
MaxPool2D
MaxPool3D
AvgPool1D
AvgPool2D
AvgPool3D
GlobalMaxPool1D
GlobalMaxPool2D
GlobalMaxPool3D
GlobalAvgPool1D
GlobalAvgPool2D
GlobalAvgPool3D
ReflectionPad2D
BatchNorm
InstanceNorm
LayerNorm
SyncBatchNorm
Embedding
SparseEmbedding
LeakyReLU
PReLU
ELU
SELU
Swish
Recurrent Layers
LSTMCell
GRUCell
RecurrentCell
SequentialRNNCell
BidirectionalCell
DropoutCell
ZoneoutCell
ResidualCell
Conv1DRNNCell
Conv2DRNNCell
Conv3DRNNCell
Conv1DLSTMCell
Conv2DLSTMCell
Conv3DLSTMCell
Conv1DGRUCell
Conv2DGRUCell
Conv3DGRUCell
VariationalDropoutCell
LSTMPCell
RNN
LSTM
GRU
Parameters
Parameter
Parameter.initialize
Parameter.data
Parameter.list_data
Parameter.list_row_sparse_data
Parameter.row_sparse_data
Parameter.set_data
Parameter.shape
Parameter.grad
Parameter.list_grad
Parameter.zero_grad
Parameter.grad_req
Parameter.cast
Parameter.list_ctx
Parameter.reset_ctx
Parameter.var
ParameterDict
ParameterDict.load
ParameterDict.save
ParameterDict.get
ParameterDict.get_constant
ParameterDict.items
ParameterDict.keys
ParameterDict.values
ParameterDict.initialize
ParameterDict.setattr
ParameterDict.update
ParameterDict.reset_ctx
ParameterDict.zero_grad
ParameterDict.prefix
mxnet.initializer
Bilinear
Constant
FusedRNN
InitDesc
Initializer
LSTMBias
Load
MSRAPrelu
Mixed
Normal
One
Orthogonal
Uniform
Xavier
Zero
register
Loss functions
Loss
L2Loss
L1Loss
SigmoidBinaryCrossEntropyLoss
SoftmaxCrossEntropyLoss
KLDivLoss
HuberLoss
HingeLoss
SquaredHingeLoss
LogisticLoss
TripletLoss
CTCLoss
Training
Trainer
Trainer.step
Trainer.allreduce_grads
Trainer.update
Trainer.load_states
Trainer.save_states
Trainer.learning_rate
Trainer.set_learning_rate
mxnet.autograd
backward
get_symbol
grad
is_recording
is_training
mark_variables
pause
predict_mode
record
set_recording
set_training
train_mode
Function
mxnet.optimizer
AdaDelta
AdaGrad
Adam
Adamax
DCASGD
FTML
Ftrl
LBSGD
NAG
Nadam
Optimizer
RMSProp
SGD
SGLD
Signum
Updater
mxnet.lr_scheduler
LRScheduler
FactorScheduler
MultiFactorScheduler
PolyScheduler
CosineScheduler
Loading Data
Dataset
ArrayDataset
RecordFileDataset
Sampler
SequentialSampler
RandomSampler
BatchSampler
DataLoader
Model Zoo
get_model
resnet18_v1
resnet34_v1
resnet50_v1
resnet101_v1
resnet152_v1
resnet18_v2
resnet34_v2
resnet50_v2
resnet101_v2
resnet152_v2
ResNetV1
ResNetV2
BasicBlockV1
BasicBlockV2
BottleneckV1
BottleneckV2
get_resnet
vgg11
vgg13
vgg16
vgg19
vgg11_bn
vgg13_bn
vgg16_bn
vgg19_bn
VGG
get_vgg
AlexNet
densenet121
densenet161
densenet169
densenet201
DenseNet
squeezenet1_0
squeezenet1_1
SqueezeNet
inception_v3
Inception3
mobilenet1_0
mobilenet0_75
mobilenet0_5
mobilenet0_25
mobilenet_v2_1_0
mobilenet_v2_0_75
mobilenet_v2_0_5
mobilenet_v2_0_25
MobileNet
MobileNetV2
Utility functions
split_data
split_and_load
clip_global_norm
download
check_sha1
Table Of Contents
Guide
Getting started with NP on MXNet
Step 1: Manipulate data with NP on MXNet
Step 2: Create a neural network
Step 3: Automatic differentiation with autograd
Step 4: Train the neural network
Step 5: Predict with a pretrained model
Step 6: Use GPUs to increase efficiency
What is NP on MXNet
API
NP on MXNet reference
Array objects
The N-dimensional array (
ndarray
)
ndarray
ndarray.shape
ndarray.ndim
ndarray.size
ndarray.dtype
ndarray.T
ndarray.item
ndarray.copy
ndarray.tolist
ndarray.astype
ndarray.reshape
ndarray.transpose
ndarray.swapaxes
ndarray.flatten
ndarray.squeeze
ndarray.nonzero
ndarray.take
ndarray.repeat
ndarray.max
ndarray.argmax
ndarray.min
ndarray.argmin
ndarray.clip
ndarray.sum
ndarray.mean
ndarray.prod
ndarray.cumsum
ndarray.var
ndarray.std
ndarray.__lt__
ndarray.__le__
ndarray.__gt__
ndarray.__ge__
ndarray.__eq__
ndarray.__ne__
ndarray.__bool__
ndarray.__neg__
ndarray.__add__
ndarray.__sub__
ndarray.__mul__
ndarray.__truediv__
ndarray.__mod__
ndarray.__pow__
ndarray.__iadd__
ndarray.__isub__
ndarray.__imul__
ndarray.__itruediv__
ndarray.__imod__
ndarray.__reduce__
ndarray.__setstate__
ndarray.__len__
ndarray.__getitem__
ndarray.__setitem__
ndarray.__int__
ndarray.__float__
ndarray.__str__
ndarray.__repr__
Indexing
Routines
Array creation routines
eye
empty
full
identity
ones
ones_like
zeros
zeros_like
array
copy
arange
linspace
logspace
meshgrid
tril
Array manipulation routines
reshape
ravel
ndarray.flatten
swapaxes
ndarray.T
transpose
moveaxis
expand_dims
squeeze
broadcast_to
broadcast_arrays
concatenate
stack
dstack
vstack
split
hsplit
vsplit
tile
repeat
unique
reshape
flip
roll
rot90
Input and output
genfromtxt
Linear algebra (
numpy.linalg
)
dot
vdot
inner
outer
tensordot
einsum
svd
norm
trace
Mathematical functions
sin
cos
tan
arcsin
arccos
arctan
degrees
radians
hypot
arctan2
deg2rad
rad2deg
sinh
cosh
tanh
arcsinh
arccosh
arctanh
rint
fix
floor
ceil
trunc
around
sum
prod
cumsum
exp
expm1
log
log10
log2
log1p
ldexp
lcm
add
reciprocal
negative
divide
power
subtract
mod
multiply
true_divide
remainder
clip
sqrt
cbrt
square
absolute
sign
maximum
minimum
np.random
choice
shuffle
normal
uniform
rand
randint
Sorting, searching, and counting
argmax
argmin
Statistics
min
max
mean
std
var
histogram
NP Extensions
set_np
reset_np
cpu
cpu_pinned
gpu
gpu_memory_info
current_context
num_gpus
activation
batch_norm
convolution
dropout
embedding
fully_connected
layer_norm
pooling
rnn
leaky_relu
multibox_detection
multibox_prior
multibox_target
roi_pooling
sigmoid
smooth_l1
softmax
mxnet.npx.threading
topk
waitall
load
save
one_hot
pick
reshape_like
batch_flatten
batch_dot
gamma
sequence_mask
Gluon reference
Creating Nerual Networks
Block objects
Block
Block.forward
Block.__call__
Block.initialize
Block.save_parameters
Block.load_parameters
Block.collect_params
Block.cast
Block.apply
Block.summary
Block.name_scope
Block.register_child
Block.register_forward_hook
Block.register_forward_pre_hook
Block.name
Block.params
Block.prefix
Block.save_params
Block.load_params
HybridBlock
HybridBlock.hybridize
HybridBlock.hybrid_forward
SymbolBlock
Container objects
Sequential
HybridSequential
Concurrent
HybridConcurrent
Feedforward Layers
Dense
Activation
Dropout
Flatten
Lambda
HybridLambda
Conv1D
Conv2D
Conv3D
Conv1DTranspose
Conv2DTranspose
Conv3DTranspose
MaxPool1D
MaxPool2D
MaxPool3D
AvgPool1D
AvgPool2D
AvgPool3D
GlobalMaxPool1D
GlobalMaxPool2D
GlobalMaxPool3D
GlobalAvgPool1D
GlobalAvgPool2D
GlobalAvgPool3D
ReflectionPad2D
BatchNorm
InstanceNorm
LayerNorm
SyncBatchNorm
Embedding
SparseEmbedding
LeakyReLU
PReLU
ELU
SELU
Swish
Recurrent Layers
LSTMCell
GRUCell
RecurrentCell
SequentialRNNCell
BidirectionalCell
DropoutCell
ZoneoutCell
ResidualCell
Conv1DRNNCell
Conv2DRNNCell
Conv3DRNNCell
Conv1DLSTMCell
Conv2DLSTMCell
Conv3DLSTMCell
Conv1DGRUCell
Conv2DGRUCell
Conv3DGRUCell
VariationalDropoutCell
LSTMPCell
RNN
LSTM
GRU
Parameters
Parameter
Parameter.initialize
Parameter.data
Parameter.list_data
Parameter.list_row_sparse_data
Parameter.row_sparse_data
Parameter.set_data
Parameter.shape
Parameter.grad
Parameter.list_grad
Parameter.zero_grad
Parameter.grad_req
Parameter.cast
Parameter.list_ctx
Parameter.reset_ctx
Parameter.var
ParameterDict
ParameterDict.load
ParameterDict.save
ParameterDict.get
ParameterDict.get_constant
ParameterDict.items
ParameterDict.keys
ParameterDict.values
ParameterDict.initialize
ParameterDict.setattr
ParameterDict.update
ParameterDict.reset_ctx
ParameterDict.zero_grad
ParameterDict.prefix
mxnet.initializer
Bilinear
Constant
FusedRNN
InitDesc
Initializer
LSTMBias
Load
MSRAPrelu
Mixed
Normal
One
Orthogonal
Uniform
Xavier
Zero
register
Loss functions
Loss
L2Loss
L1Loss
SigmoidBinaryCrossEntropyLoss
SoftmaxCrossEntropyLoss
KLDivLoss
HuberLoss
HingeLoss
SquaredHingeLoss
LogisticLoss
TripletLoss
CTCLoss
Training
Trainer
Trainer.step
Trainer.allreduce_grads
Trainer.update
Trainer.load_states
Trainer.save_states
Trainer.learning_rate
Trainer.set_learning_rate
mxnet.autograd
backward
get_symbol
grad
is_recording
is_training
mark_variables
pause
predict_mode
record
set_recording
set_training
train_mode
Function
mxnet.optimizer
AdaDelta
AdaGrad
Adam
Adamax
DCASGD
FTML
Ftrl
LBSGD
NAG
Nadam
Optimizer
RMSProp
SGD
SGLD
Signum
Updater
mxnet.lr_scheduler
LRScheduler
FactorScheduler
MultiFactorScheduler
PolyScheduler
CosineScheduler
Loading Data
Dataset
ArrayDataset
RecordFileDataset
Sampler
SequentialSampler
RandomSampler
BatchSampler
DataLoader
Model Zoo
get_model
resnet18_v1
resnet34_v1
resnet50_v1
resnet101_v1
resnet152_v1
resnet18_v2
resnet34_v2
resnet50_v2
resnet101_v2
resnet152_v2
ResNetV1
ResNetV2
BasicBlockV1
BasicBlockV2
BottleneckV1
BottleneckV2
get_resnet
vgg11
vgg13
vgg16
vgg19
vgg11_bn
vgg13_bn
vgg16_bn
vgg19_bn
VGG
get_vgg
AlexNet
densenet121
densenet161
densenet169
densenet201
DenseNet
squeezenet1_0
squeezenet1_1
SqueezeNet
inception_v3
Inception3
mobilenet1_0
mobilenet0_75
mobilenet0_5
mobilenet0_25
mobilenet_v2_1_0
mobilenet_v2_0_75
mobilenet_v2_0_5
mobilenet_v2_0_25
MobileNet
MobileNetV2
Utility functions
split_data
split_and_load
clip_global_norm
download
check_sha1
Gluon
ΒΆ
TODO
