megengine.functional.nn.embedding

embedding(inp, weight, padding_idx=None, max_norm=None, norm_type=None)

Applies lookup table for embedding.

Parameters

• inp (Tensor) – tensor with indices.

• weight (Tensor) – learnable weights which embeds from.

• padding_idx (Optional[int]) – should be set to None, not supported now.

• max_norm (Optional[float]) – should be set to None, not supported now.

• norm_type (Optional[float]) – should be set to None, not supported now.

Refer to Embedding for more information.

previous

megengine.functional.nn.indexing_one_hot

next

megengine.functional.nn.l1_loss