timesead.models.layers.autoformer_encdec

Classes

`CustomLayerNorm`	Special designed layernorm for the seasonal part
`MovingAvg`	Moving average block to highlight the trend of time series
`SeriesDecomp`	Series decomposition block
`EncoderLayer`	Autoformer encoder layer with the progressive decomposition architecture
`Encoder`	Autoformer encoder

Module Contents

class timesead.models.layers.autoformer_encdec.CustomLayerNorm(channels)

Bases: torch.nn.Module

Special designed layernorm for the seasonal part

Initialize internal Module state, shared by both nn.Module and ScriptModule.

layernorm

forward(x)

class timesead.models.layers.autoformer_encdec.MovingAvg(kernel_size, stride)

Bases: torch.nn.Module

Moving average block to highlight the trend of time series

Initialize internal Module state, shared by both nn.Module and ScriptModule.

kernel_size

avg

forward(x)

class timesead.models.layers.autoformer_encdec.SeriesDecomp(kernel_size)

Bases: torch.nn.Module

Series decomposition block

Initialize internal Module state, shared by both nn.Module and ScriptModule.

moving_avg

forward(x)

class timesead.models.layers.autoformer_encdec.EncoderLayer(attention, d_model, d_ff=None, moving_avg=25, dropout=0.1, activation='relu')

Bases: torch.nn.Module

Autoformer encoder layer with the progressive decomposition architecture

Initialize internal Module state, shared by both nn.Module and ScriptModule.

attention

conv1

conv2

decomp1

decomp2

dropout

activation

forward(x, attn_mask=None)

class timesead.models.layers.autoformer_encdec.Encoder(attn_layers, conv_layers=None, norm_layer=None)

Bases: torch.nn.Module

Autoformer encoder

Initialize internal Module state, shared by both nn.Module and ScriptModule.

attn_layers

conv_layers

norm = None

forward(x, attn_mask=None)