Loading...
Searching...
No Matches
standard_networks.h
Go to the documentation of this file.
ApproximationNetwork(const Shape &input_shape, const Shape &complexity_dimensions, const Shape &output_shape)
Builds an approximation network with the given input, complexity and output shapes.
AutoAssociationNetwork(const Shape &input_shape, const Shape &complexity_dimensions, const Shape &output_shape)
Builds an auto-association network with the given input, complexity and output shapes.
ClassificationNetwork(const Shape &input_shape, const Shape &complexity_dimensions, const Shape &output_shape)
Builds a classification network with the given input, complexity and output shapes.
ForecastingNetwork(const Shape &input_shape, const Shape &complexity_dimensions, const Shape &output_shape)
Builds a forecasting network with the given input, complexity and output shapes.
ImageClassificationNetwork(const Shape &input_shape, const Shape &complexity_dimensions, const Shape &output_shape)
Builds an image classification network with the given input, complexity and output shapes.
SimpleResNet(const Shape &input_shape, const vector< Index > &blocks_per_stage, const Shape &initial_filters, const Shape &output_shape)
Builds a residual network with the given input shape, per-stage block counts and output shape.
TextClassificationNetwork(const Shape &input_shape, const Shape &complexity_dimensions, const Shape &output_shape)
Builds a text classification network with the given input, complexity and output shapes.
Index get_heads_number() const
Returns the number of attention heads per block.
Index get_decoder_sequence_length() const
Returns the configured length of the decoder sequence.
void set_dropout_rate(const float)
Sets the dropout rate applied across the Transformer layers.
Transformer(const Index=0, Index=0, Index=0, Index=0, Index=0, Index=0, Index=0, Index=0)
Builds a Transformer with input/decoder sequence lengths, vocabularies, embedding,...
Index get_embedding_dimension() const
Returns the embedding dimension used by the Transformer.
void set(const Index=0, Index=0, Index=0, Index=0, Index=0, Index=0, Index=0, Index=0)
Reconfigures the Transformer with the supplied architecture parameters.
Index get_input_sequence_length() const
Returns the configured length of the encoder input sequence.
VGG16(const Shape &input_shape, const Shape &target_shape)
Builds a VGG-16 network with the given input and target shapes.
void set(const Shape &input_shape, const Shape &target_shape)
Reconfigures the VGG-16 network with the given input and target shapes.
VGG16(const filesystem::path &)
Builds a VGG-16 network by loading it from the given file path.
Definition adaptive_moment_estimation.h:14
Fixed-capacity small-vector describing tensor dimensions (rank up to MaxRank).
Definition tensor_utilities.h:42