patchify_id_mask

embedding.patchify_id_mask

(No summary available)

Parameter	Type	Default	Description
id_mask	Int[torch.Tensor, "batch variate time_steps"]	-	(undocumented)
patch_size	int	-	(undocumented)

Returns: Int[torch.Tensor, "batch variate seq_len patch_size"] (undocumented)

PatchEmbedding

embedding.PatchEmbedding · inherits torch.nn.Module

Multivariate time series patch embedding. Patchifies each variate separately.

__init__(self, patch_size: int, stride: int, embed_dim: int)

(No prose summary)

Parameter	Type	Default	Description
patch_size	int	-	(undocumented)
stride	int	-	(undocumented)
embed_dim	int	-	(undocumented)

forward(self, x, id_mask)

(No prose summary)

Parameter	Type	Default	Description
x	Float[torch.Tensor, "batch #variate time_steps"]	-	(undocumented)
id_mask	Float[torch.Tensor, "batch time_steps"]	-	(undocumented)

Returns: tuple[ Float[torch.Tensor, "batch variate seq_len embed_dim"], Int[torch.Tensor, "batch seq_len"], ] (undocumented)