TODO.md

Encoders

Docs

Binary

• Passthrough -> passenc
• Dense -> denc

Numbers

• Passthrough
• Dense

Categorical

• Passthrough
• Sparse

Bag

• EmbedWeighted

Set

• Embed

Sequence

• Embed
• Parallel CNN
• Stacked CNN
• Stacked Parallel CNN
• RNN
• CNN RNN
• Transformer

Text

• Embed
• Parallel CNN
• Stacked CNN
• Stacked Parallel CNN
• RNN
• CNN RNN
• Transformer
• Huggingface

Vector

• PassThrough -> passenc
• Dense -> denc

Audio

(Same like in Sequence)

• Embed
• Parallel CNN
• Stacked CNN
• Stacked Parallel CNN
• RNN
• CNN RNN
• Transformer

Date

• Embed
• Wave

H3

• Embed
• Weighted Sum Embed
• RNN

Image

• Convolutional Stack
• MLP-Mixer
• Torchvision
• U-Net

Time Series

(Same like in Sequence)

• Embed
• Parallel CNN
• Stacked CNN
• Stacked Parallel CNN
• RNN
• CNN RNN
• Transformer

Hat tatsächlich Inhalt.

• Passthrough -> tspassenc

Files

encoders/base.py

• PassthroughEncoder -> passenc
• DenseEncoder -> denc

encoders/bag_encoders

• BagEmbedWeightedEncoder

encoders/category_encoders

• CategoricalEmbedEncoder
• CategoricalOneHotEncoder (Seems not being used)
• CategoricalPassthroughEncoder (in base.py)
• CategoricalSparseEncoder (Dense)

encoders/date_encoders

• DateEmbed
• DateWave

encoders/h3_encoders

• H3EmbedConfig
• H3RNNConfig
• H3WeightedSumConfig

encoders/sequence_encoders

• SequenceEmbedConfig
• SequenceParallelCNNConfig
• SequenceStackedCNNConfig
• SequenceStackedParallelCNNConfig
• SequenceRNNConfig
• SequenceCNNRNNConfig
• SequenceTransformerConfig

encoders/set_encoders

• SetSparseEncoderConfig

encoders/image/base

• ImageEncoderConfig

encoders/image/torchvision

• TVBaseEncoderConfig