Deep Learning with 'mlr3'

Documentation for package ‘mlr3torch’ version 0.4.0

Help Pages

mlr3torch: Deep Learning with 'mlr3'

-- A --

assert_dim_in_range	Assert that a Dimension Exists
assert_known_dims	Assert that Dimensions are Known
assert_lazy_tensor	Assert Lazy Tensor
assert_ndim	Assert the Number of Dimensions of a Shape
assert_not_batch_dim	Assert that a Dimension is not the Batch Dimension
assert_positive_extent	Assert that a Computed Output Size is Positive
assert_same_batch_size	Assert that Shapes have the Same Batch Size
assert_same_ndim	Assert that Shapes have the Same Number of Dimensions
assert_shape	Assert a Shape
assert_shapes	Assert a List of Shapes
as_data_descriptor	Convert to Data Descriptor
as_lazy_tensor	Convert to Lazy Tensor
as_lazy_tensor.dataset	Convert to Lazy Tensor
as_learner_torch	Convert a Graph to a Torch Learner
as_learner_torch.Graph	Convert a Graph to a Torch Learner
as_learner_torch.GraphLearner	Convert a Graph to a Torch Learner
as_learner_torch.PipeOp	Convert a Graph to a Torch Learner
as_lr_scheduler	Convert to CallbackSetLRScheduler
as_task_torch	Create a Generic Torch Task
as_torch_callback	Convert to a TorchCallback
as_torch_callbacks	Convert to a list of Torch Callbacks
as_torch_loss	Convert to TorchLoss
as_torch_optimizer	Convert to TorchOptimizer
auto_device	Auto Device

-- B --

batchgetter_categ	Batchgetter for Categorical data
batchgetter_num	Batchgetter for Numeric Data
broadcast_shapes	Helpers for Shape Inference

-- C --

CallbackSet	Base Class for Callbacks
CallbackSetCheckpoint	Checkpoint Callback
CallbackSetHistory	History Callback
CallbackSetLRScheduler	Learning Rate Scheduling Callback
CallbackSetLRSchedulerOneCycle	OneCycle Learning Rate Scheduling Callback
CallbackSetLRSchedulerReduceOnPlateau	Reduce On Plateau Learning Rate Scheduler
CallbackSetProgress	Progress Callback
CallbackSetTB	TensorBoard Logging Callback
CallbackSetUnfreeze	Unfreezing Weights Callback
callback_set	Create a Set of Callbacks for Torch
ContextTorch	Context for Torch Learner
cross_entropy	Cross Entropy Loss

-- D --

DataBackendLazy	Lazy Data Backend
DataDescriptor	Data Descriptor

-- E --

encode_prediction

Encode the Network Output as a Prediction

-- G --

get_batch_constructor	Batch Constructor for a Task
get_batch_size	Extract the Batch Size for a Given Phase
get_target_batchgetter	Target Batchgetter for a Task
GraphLearnerTorch	Graph Learner for Torch Networks

-- I --

infer_shapes	Infer Shapes
ingress_categ	Ingress Token for Categorical Features
ingress_ltnsr	Ingress Token for Lazy Tensor Feature
ingress_num	Ingress Token for Numeric Features
is_lazy_tensor	Check for lazy tensor

-- L --

lazy_shape	Shape of Lazy Tensor
lazy_tensor	Create a lazy tensor
LearnerTorch	Base Class for Torch Learners
LearnerTorchFeatureless	Featureless Torch Learner
LearnerTorchFTTransformer	FT-Transformer
LearnerTorchImage	Image Learner
LearnerTorchMLP	Multi Layer Perceptron
LearnerTorchModel	Learner Torch Model
LearnerTorchModule	Learner Torch Module
LearnerTorchRealMLP	RealMLP
LearnerTorchTabM	TabM
LearnerTorchTabResNet	Tabular ResNet
LearnerTorchVision	Image Classification Networks from torchvision

-- M --

materialize	Materialize Lazy Tensor Columns
materialize.list	Materialize Lazy Tensor Columns
MeasureTorch	Measure for a Generic Torch Task
MeasureTorchDefault	Default Measure of a Generic Torch Task
mlr3torch	mlr3torch: Deep Learning with 'mlr3'
mlr3torch_callbacks	Dictionary of Torch Callbacks
mlr3torch_losses	Loss Functions
mlr3torch_optimizers	Optimizers
mlr_backends_lazy	Lazy Data Backend
mlr_callback_set	Base Class for Callbacks
mlr_callback_set.checkpoint	Checkpoint Callback
mlr_callback_set.history	History Callback
mlr_callback_set.lr_scheduler	Learning Rate Scheduling Callback
mlr_callback_set.lr_scheduler_one_cycle	OneCycle Learning Rate Scheduling Callback
mlr_callback_set.lr_scheduler_reduce_on_plateau	Reduce On Plateau Learning Rate Scheduler
mlr_callback_set.progress	Progress Callback
mlr_callback_set.tb	TensorBoard Logging Callback
mlr_callback_set.unfreeze	Unfreezing Weights Callback
mlr_context_torch	Context for Torch Learner
mlr_learners.ft_transformer	FT-Transformer
mlr_learners.mlp	Multi Layer Perceptron
mlr_learners.module	Learner Torch Module
mlr_learners.realmlp	RealMLP
mlr_learners.tabm	TabM
mlr_learners.tab_resnet	Tabular ResNet
mlr_learners.torchvision	Image Classification Networks from torchvision
mlr_learners.torch_featureless	Featureless Torch Learner
mlr_learners_torch	Base Class for Torch Learners
mlr_learners_torch_image	Image Learner
mlr_learners_torch_model	Learner Torch Model
mlr_measures_torch	Measure for a Generic Torch Task
mlr_measures_torch.default	Default Measure of a Generic Torch Task
mlr_pipeops_augment_center_crop	Center Crop Augmentation
mlr_pipeops_augment_color_jitter	Color Jitter Augmentation
mlr_pipeops_augment_crop	Crop Augmentation
mlr_pipeops_augment_hflip	Horizontal Flip Augmentation
mlr_pipeops_augment_random_affine	Random Affine Augmentation
mlr_pipeops_augment_random_choice	Random Choice Augmentation
mlr_pipeops_augment_random_crop	Random Crop Augmentation
mlr_pipeops_augment_random_horizontal_flip	Random Horizontal Flip Augmentation
mlr_pipeops_augment_random_order	Random Order Augmentation
mlr_pipeops_augment_random_resized_crop	Random Resized Crop Augmentation
mlr_pipeops_augment_random_vertical_flip	Random Vertical Flip Augmentation
mlr_pipeops_augment_resized_crop	Resized Crop Augmentation
mlr_pipeops_augment_rotate	Rotate Augmentation
mlr_pipeops_augment_vflip	Vertical Flip Augmentation
mlr_pipeops_module	Class for Torch Module Wrappers
mlr_pipeops_nn_adaptive_avg_pool1d	1D Adaptive Average Pooling
mlr_pipeops_nn_adaptive_avg_pool2d	2D Adaptive Average Pooling
mlr_pipeops_nn_adaptive_avg_pool3d	3D Adaptive Average Pooling
mlr_pipeops_nn_adaptive_max_pool1d	1D Adaptive Max Pooling
mlr_pipeops_nn_adaptive_max_pool2d	2D Adaptive Max Pooling
mlr_pipeops_nn_adaptive_max_pool3d	3D Adaptive Max Pooling
mlr_pipeops_nn_avg_pool1d	1D Average Pooling
mlr_pipeops_nn_avg_pool2d	2D Average Pooling
mlr_pipeops_nn_avg_pool3d	3D Average Pooling
mlr_pipeops_nn_batch_norm1d	1D Batch Normalization
mlr_pipeops_nn_batch_norm2d	2D Batch Normalization
mlr_pipeops_nn_batch_norm3d	3D Batch Normalization
mlr_pipeops_nn_block	Block Repetition
mlr_pipeops_nn_celu	CELU Activation Function
mlr_pipeops_nn_conv1d	1D Convolution
mlr_pipeops_nn_conv2d	2D Convolution
mlr_pipeops_nn_conv3d	3D Convolution
mlr_pipeops_nn_conv_transpose1d	Transpose 1D Convolution
mlr_pipeops_nn_conv_transpose2d	Transpose 2D Convolution
mlr_pipeops_nn_conv_transpose3d	Transpose 3D Convolution
mlr_pipeops_nn_dropout	Dropout
mlr_pipeops_nn_dropout2d	2D Dropout
mlr_pipeops_nn_dropout3d	3D Dropout
mlr_pipeops_nn_elu	ELU Activation Function
mlr_pipeops_nn_flatten	Flattens a Tensor
mlr_pipeops_nn_fn	Custom Function
mlr_pipeops_nn_ft_cls	CLS Token for FT-Transformer
mlr_pipeops_nn_ft_transformer_block	Single Transformer Block for the FT-Transformer
mlr_pipeops_nn_geglu	GeGLU Activation Function
mlr_pipeops_nn_gelu	GELU Activation Function
mlr_pipeops_nn_glu	GLU Activation Function
mlr_pipeops_nn_group_norm	Group Normalization
mlr_pipeops_nn_gru	Gated Recurrent Unit
mlr_pipeops_nn_hardshrink	Hard Shrink Activation Function
mlr_pipeops_nn_hardsigmoid	Hard Sigmoid Activation Function
mlr_pipeops_nn_hardtanh	Hard Tanh Activation Function
mlr_pipeops_nn_head	Output Head
mlr_pipeops_nn_identity	Identity Layer
mlr_pipeops_nn_layer_norm	Layer Normalization
mlr_pipeops_nn_leaky_relu	Leaky ReLU Activation Function
mlr_pipeops_nn_linear	Linear Layer
mlr_pipeops_nn_log_sigmoid	Log Sigmoid Activation Function
mlr_pipeops_nn_log_softmax	Log Softmax
mlr_pipeops_nn_lstm	Long Short-Term Memory
mlr_pipeops_nn_max_pool1d	1D Max Pooling
mlr_pipeops_nn_max_pool2d	2D Max Pooling
mlr_pipeops_nn_max_pool3d	3D Max Pooling
mlr_pipeops_nn_merge	Merge Operation
mlr_pipeops_nn_merge_cat	Merge by Concatenation
mlr_pipeops_nn_merge_prod	Merge by Product
mlr_pipeops_nn_merge_sum	Merge by Summation
mlr_pipeops_nn_multihead_attention	Multi-Head Attention
mlr_pipeops_nn_prelu	PReLU Activation Function
mlr_pipeops_nn_recurrent	Recurrent Layer
mlr_pipeops_nn_reglu	ReGLU Activation Function
mlr_pipeops_nn_relu	ReLU Activation Function
mlr_pipeops_nn_relu6	ReLU6 Activation Function
mlr_pipeops_nn_reshape	Reshape a Tensor
mlr_pipeops_nn_rnn	Simple Recurrent Layer
mlr_pipeops_nn_rrelu	RReLU Activation Function
mlr_pipeops_nn_selu	SELU Activation Function
mlr_pipeops_nn_sigmoid	Sigmoid Activation Function
mlr_pipeops_nn_silu	SiLU Activation Function
mlr_pipeops_nn_softmax	Softmax
mlr_pipeops_nn_softmax2d	2D Softmax
mlr_pipeops_nn_softmin	Softmin
mlr_pipeops_nn_softplus	SoftPlus Activation Function
mlr_pipeops_nn_softshrink	Soft Shrink Activation Function
mlr_pipeops_nn_softsign	SoftSign Activation Function
mlr_pipeops_nn_squeeze	Squeeze a Tensor
mlr_pipeops_nn_tanh	Tanh Activation Function
mlr_pipeops_nn_tanhshrink	Tanh Shrink Activation Function
mlr_pipeops_nn_threshold	Threshold Activation Function
mlr_pipeops_nn_tokenizer_categ	Categorical Tokenizer
mlr_pipeops_nn_tokenizer_num	Numeric Tokenizer
mlr_pipeops_nn_transformer_encoder_layer	Transformer Encoder Layer
mlr_pipeops_nn_unflatten	Unflattens a Tensor
mlr_pipeops_nn_unsqueeze	Unsqueeze a Tensor
mlr_pipeops_preproc_torch	Base Class for Lazy Tensor Preprocessing
mlr_pipeops_torch	Base Class for Torch Module Constructor Wrappers
mlr_pipeops_torch_callbacks	Callback Configuration
mlr_pipeops_torch_ingress	Entrypoint to Torch Network
mlr_pipeops_torch_ingress_categ	Torch Entry Point for Categorical Features
mlr_pipeops_torch_ingress_ltnsr	Ingress for Lazy Tensor
mlr_pipeops_torch_ingress_num	Torch Entry Point for Numeric Features
mlr_pipeops_torch_loss	Loss Configuration
mlr_pipeops_torch_model	PipeOp Torch Model
mlr_pipeops_torch_model_classif	PipeOp Torch Classifier
mlr_pipeops_torch_model_regr	Torch Regression Model
mlr_pipeops_torch_optimizer	Optimizer Configuration
mlr_pipeops_trafo_adjust_brightness	Adjust Brightness Transformation
mlr_pipeops_trafo_adjust_gamma	Adjust Gamma Transformation
mlr_pipeops_trafo_adjust_hue	Adjust Hue Transformation
mlr_pipeops_trafo_adjust_saturation	Adjust Saturation Transformation
mlr_pipeops_trafo_grayscale	Grayscale Transformation
mlr_pipeops_trafo_nop	No Transformation
mlr_pipeops_trafo_normalize	Normalization Transformation
mlr_pipeops_trafo_pad	Padding Transformation
mlr_pipeops_trafo_reshape	Reshaping Transformation
mlr_pipeops_trafo_resize	Resizing Transformation
mlr_pipeops_trafo_rgb_to_grayscale	RGB to Grayscale Transformation
mlr_tasks_cifar	CIFAR Classification Tasks
mlr_tasks_cifar10	CIFAR Classification Tasks
mlr_tasks_cifar100	CIFAR Classification Tasks
mlr_tasks_lazy_iris	Iris Classification Task
mlr_tasks_melanoma	Melanoma Image classification
mlr_tasks_mnist	MNIST Image classification
mlr_tasks_tiny_imagenet	Tiny ImageNet Classification Task
mlr_tasks_torch	Generic Torch Task
ModelDescriptor	Represent a Model with Meta-Info
model_descriptor_to_learner	Create a Torch Learner from a ModelDescriptor
model_descriptor_to_module	Create a nn_graph from ModelDescriptor
model_descriptor_union	Union of ModelDescriptors
msr_torch	Create a Measure for a Generic Torch Task

-- N --

nn	Create a Neural Network Layer
nn_ft_cls	CLS Token for FT-Transformer
nn_ft_transformer_block	Single Transformer Block for FT-Transformer
nn_geglu	GeGLU Module
nn_graph	Graph Network
nn_merge_cat	Concatenates multiple tensors
nn_merge_prod	Product of multiple tensors
nn_merge_sum	Sum of multiple tensors
nn_reglu	ReGLU Module
nn_reshape	Reshape
nn_squeeze	Squeeze
nn_tokenizer_categ	Categorical Tokenizer
nn_tokenizer_num	Numeric Tokenizer
nn_unsqueeze	Unsqueeze

-- O --

output_dim_for

Network Output Dimension

-- P --

PipeOpModule	Class for Torch Module Wrappers
PipeOpPreprocTorchAugmentCenterCrop	Center Crop Augmentation
PipeOpPreprocTorchAugmentColorJitter	Color Jitter Augmentation
PipeOpPreprocTorchAugmentCrop	Crop Augmentation
PipeOpPreprocTorchAugmentHflip	Horizontal Flip Augmentation
PipeOpPreprocTorchAugmentRandomAffine	Random Affine Augmentation
PipeOpPreprocTorchAugmentRandomChoice	Random Choice Augmentation
PipeOpPreprocTorchAugmentRandomCrop	Random Crop Augmentation
PipeOpPreprocTorchAugmentRandomHorizontalFlip	Random Horizontal Flip Augmentation
PipeOpPreprocTorchAugmentRandomOrder	Random Order Augmentation
PipeOpPreprocTorchAugmentRandomResizedCrop	Random Resized Crop Augmentation
PipeOpPreprocTorchAugmentRandomVerticalFlip	Random Vertical Flip Augmentation
PipeOpPreprocTorchAugmentResizedCrop	Resized Crop Augmentation
PipeOpPreprocTorchAugmentRotate	Rotate Augmentation
PipeOpPreprocTorchAugmentVflip	Vertical Flip Augmentation
PipeOpPreprocTorchTrafoAdjustBrightness	Adjust Brightness Transformation
PipeOpPreprocTorchTrafoAdjustGamma	Adjust Gamma Transformation
PipeOpPreprocTorchTrafoAdjustHue	Adjust Hue Transformation
PipeOpPreprocTorchTrafoAdjustSaturation	Adjust Saturation Transformation
PipeOpPreprocTorchTrafoGrayscale	Grayscale Transformation
PipeOpPreprocTorchTrafoNop	No Transformation
PipeOpPreprocTorchTrafoNormalize	Normalization Transformation
PipeOpPreprocTorchTrafoPad	Padding Transformation
PipeOpPreprocTorchTrafoReshape	Reshaping Transformation
PipeOpPreprocTorchTrafoResize	Resizing Transformation
PipeOpPreprocTorchTrafoRgbToGrayscale	RGB to Grayscale Transformation
PipeOpTaskPreprocTorch	Base Class for Lazy Tensor Preprocessing
PipeOpTorch	Base Class for Torch Module Constructor Wrappers
PipeOpTorchAdaptiveAvgPool1D	1D Adaptive Average Pooling
PipeOpTorchAdaptiveAvgPool2D	2D Adaptive Average Pooling
PipeOpTorchAdaptiveAvgPool3D	3D Adaptive Average Pooling
PipeOpTorchAdaptiveMaxPool1D	1D Adaptive Max Pooling
PipeOpTorchAdaptiveMaxPool2D	2D Adaptive Max Pooling
PipeOpTorchAdaptiveMaxPool3D	3D Adaptive Max Pooling
PipeOpTorchAvgPool1D	1D Average Pooling
PipeOpTorchAvgPool2D	2D Average Pooling
PipeOpTorchAvgPool3D	3D Average Pooling
PipeOpTorchBatchNorm1D	1D Batch Normalization
PipeOpTorchBatchNorm2D	2D Batch Normalization
PipeOpTorchBatchNorm3D	3D Batch Normalization
PipeOpTorchBlock	Block Repetition
PipeOpTorchCallbacks	Callback Configuration
PipeOpTorchCELU	CELU Activation Function
PipeOpTorchConv1D	1D Convolution
PipeOpTorchConv2D	2D Convolution
PipeOpTorchConv3D	3D Convolution
PipeOpTorchConvTranspose1D	Transpose 1D Convolution
PipeOpTorchConvTranspose2D	Transpose 2D Convolution
PipeOpTorchConvTranspose3D	Transpose 3D Convolution
PipeOpTorchDropout	Dropout
PipeOpTorchDropout2D	2D Dropout
PipeOpTorchDropout3D	3D Dropout
PipeOpTorchELU	ELU Activation Function
PipeOpTorchFlatten	Flattens a Tensor
PipeOpTorchFn	Custom Function
PipeOpTorchFTCLS	CLS Token for FT-Transformer
PipeOpTorchFTTransformerBlock	Single Transformer Block for the FT-Transformer
PipeOpTorchGeGLU	GeGLU Activation Function
PipeOpTorchGELU	GELU Activation Function
PipeOpTorchGLU	GLU Activation Function
PipeOpTorchGroupNorm	Group Normalization
PipeOpTorchGRU	Gated Recurrent Unit
PipeOpTorchHardShrink	Hard Shrink Activation Function
PipeOpTorchHardSigmoid	Hard Sigmoid Activation Function
PipeOpTorchHardTanh	Hard Tanh Activation Function
PipeOpTorchHead	Output Head
PipeOpTorchIdentity	Identity Layer
PipeOpTorchIngress	Entrypoint to Torch Network
PipeOpTorchIngressCategorical	Torch Entry Point for Categorical Features
PipeOpTorchIngressLazyTensor	Ingress for Lazy Tensor
PipeOpTorchIngressNumeric	Torch Entry Point for Numeric Features
PipeOpTorchLayerNorm	Layer Normalization
PipeOpTorchLeakyReLU	Leaky ReLU Activation Function
PipeOpTorchLinear	Linear Layer
PipeOpTorchLogSigmoid	Log Sigmoid Activation Function
PipeOpTorchLogSoftmax	Log Softmax
PipeOpTorchLoss	Loss Configuration
PipeOpTorchLSTM	Long Short-Term Memory
PipeOpTorchMaxPool1D	1D Max Pooling
PipeOpTorchMaxPool2D	2D Max Pooling
PipeOpTorchMaxPool3D	3D Max Pooling
PipeOpTorchMerge	Merge Operation
PipeOpTorchMergeCat	Merge by Concatenation
PipeOpTorchMergeProd	Merge by Product
PipeOpTorchMergeSum	Merge by Summation
PipeOpTorchModel	PipeOp Torch Model
PipeOpTorchModelClassif	PipeOp Torch Classifier
PipeOpTorchModelRegr	Torch Regression Model
PipeOpTorchMultiheadAttention	Multi-Head Attention
PipeOpTorchOptimizer	Optimizer Configuration
PipeOpTorchPReLU	PReLU Activation Function
PipeOpTorchRecurrent	Recurrent Layer
PipeOpTorchReGLU	ReGLU Activation Function
PipeOpTorchReLU	ReLU Activation Function
PipeOpTorchReLU6	ReLU6 Activation Function
PipeOpTorchReshape	Reshape a Tensor
PipeOpTorchRNN	Simple Recurrent Layer
PipeOpTorchRReLU	RReLU Activation Function
PipeOpTorchSELU	SELU Activation Function
PipeOpTorchSigmoid	Sigmoid Activation Function
PipeOpTorchSiLU	SiLU Activation Function
PipeOpTorchSoftmax	Softmax
PipeOpTorchSoftmax2D	2D Softmax
PipeOpTorchSoftmin	Softmin
PipeOpTorchSoftPlus	SoftPlus Activation Function
PipeOpTorchSoftShrink	Soft Shrink Activation Function
PipeOpTorchSoftSign	SoftSign Activation Function
PipeOpTorchSqueeze	Squeeze a Tensor
PipeOpTorchTanh	Tanh Activation Function
PipeOpTorchTanhShrink	Tanh Shrink Activation Function
PipeOpTorchThreshold	Threshold Activation Function
PipeOpTorchTokenizerCateg	Categorical Tokenizer
PipeOpTorchTokenizerNum	Numeric Tokenizer
PipeOpTorchTransformerEncoderLayer	Transformer Encoder Layer
PipeOpTorchUnflatten	Unflattens a Tensor
PipeOpTorchUnsqueeze	Unsqueeze a Tensor
pipeop_preproc_torch	Create Torch Preprocessing PipeOps
pipeop_torch	Create a PipeOpTorch
PredictionTorch	Prediction Object for a Generic Torch Task

-- R --

reshape_output_shape	Output Shape of a Reshape
resolve_dim	Helpers for Shape Inference

-- S --

Select	Selector Functions for Character Vectors
select_all	Selector Functions for Character Vectors
select_grep	Selector Functions for Character Vectors
select_invert	Selector Functions for Character Vectors
select_name	Selector Functions for Character Vectors
select_none	Selector Functions for Character Vectors
shape_helpers	Helpers for Shape Inference
shape_to_str	Helpers for Shape Inference

-- T --

TaskTorch	Generic Torch Task
task_dataset	Create a Dataset from a Task
TorchCallback	Torch Callback
TorchDescriptor	Base Class for Torch Descriptors
TorchIngressToken	Torch Ingress Token
TorchLoss	Torch Loss
TorchOptimizer	Torch Optimizer
torch_callback	Create a Callback Descriptor
t_clbk	Sugar Function for Torch Callback
t_clbks	Sugar Function for Torch Callback
t_loss	Loss Function Quick Access
t_losses	Loss Function Quick Access
t_opt	Optimizers Quick Access
t_opts	Optimizers Quick Access