attention_utils

class Linear3D(hidden_size, num_attention_heads, size_per_head, weight_attr=None, bias_attr=None)

forward(input)

Defines the computation performed at every call. Should be overridden by all subclasses.

Parameters

• *inputs (tuple) – unpacked tuple arguments

• **kwargs (dict) – unpacked dict arguments

class Attention(num_heads=1, block_size=1, window_size=3, num_global_blocks=1, num_rand_blocks=1, seed=None)

forward(query_matrix, key_matrix, value_matrix, d_head, attn_mask=None, rand_mask_idx=None, query_mask=None, key_mask=None, dropout=None)

Defines the computation performed at every call. Should be overridden by all subclasses.

Parameters

• *inputs (tuple) – unpacked tuple arguments

• **kwargs (dict) – unpacked dict arguments

class DefaultAttention(num_heads=1, block_size=1, window_size=3, num_global_blocks=1, num_rand_blocks=1, seed=None)

forward(query_matrix, key_matrix, value_matrix, d_head, attn_mask=None, rand_mask_idx=None, query_mask=None, key_mask=None, dropout=None)

Defines the computation performed at every call. Should be overridden by all subclasses.

Parameters

• *inputs (tuple) – unpacked tuple arguments

• **kwargs (dict) – unpacked dict arguments

class BigBirdSparseAttention(num_heads=1, block_size=1, window_size=3, num_global_blocks=1, num_rand_blocks=1, seed=None)

forward(query_matrix, key_matrix, value_matrix, d_head, attn_mask=None, rand_mask_idx=None, query_mask=None, key_mask=None, dropout=None)

query_matrix: [B, H, T, D] key_matrix: [B, H, T, D] value_matrix: [B, H, T, D] query_mask: [B, 1, T, 1] bool mask key_mask: [B, 1, 1, T] bool mask rand_mask_idx: [H, T//bs, bs] Global Attention Random Attention Window Attention

class MultiHeadAttention(embed_dim, num_heads, dropout=0.0, kdim=None, vdim=None, weight_attr=None, bias_attr=None, block_size=1, window_size=3, num_global_blocks=1, num_rand_blocks=1, seed=None, attention_type='bigbird')

class Cache(k, v)

k

Alias for field number 0

v

Alias for field number 1

class StaticCache(k, v)

k

Alias for field number 0

v

Alias for field number 1

forward(query, key, value, attn_mask=None, rand_mask_idx=None, query_mask=None, key_mask=None, cache=None)

Defines the computation performed at every call. Should be overridden by all subclasses.

Parameters

• *inputs (tuple) – unpacked tuple arguments

• **kwargs (dict) – unpacked dict arguments