ItemNetBase

class rectools.models.nn.item_net.ItemNetBase(*args: Any, **kwargs: Any)

Bases: Module

Base class for item net.

Methods

forward(items)	Forward pass.
from_dataset(dataset, *args, **kwargs)	Construct ItemNet from Dataset.
from_dataset_schema(dataset_schema, *args, ...)	Construct ItemNet from Dataset schema.
get_all_embeddings()	Return item embeddings.

Attributes

device	Return ItemNet device.
out_dim	Return item embedding output dimension.

Parameters

• args (Any) –

• kwargs (Any) –

property device: device

Return ItemNet device.

forward(items: Tensor) → Tensor

Forward pass.

Parameters

items (Tensor) –

Return type

Tensor

classmethod from_dataset(dataset: Dataset, *args: Any, **kwargs: Any) → Optional[Self]

Construct ItemNet from Dataset.

Parameters

• dataset (Dataset) –

• args (Any) –

• kwargs (Any) –

Return type

Optional[Self]

classmethod from_dataset_schema(dataset_schema: DatasetSchema, *args: Any, **kwargs: Any) → Optional[Self]

Construct ItemNet from Dataset schema.

Parameters

• dataset_schema (DatasetSchema) –

• args (Any) –

• kwargs (Any) –

Return type

Optional[Self]

get_all_embeddings() → Tensor

Return item embeddings.

Return type

Tensor

property out_dim: int

Return item embedding output dimension.