Source code for gluonnlp.model.attention_cell

# coding: utf-8

# Licensed to the Apache Software Foundation (ASF) under one
# or more contributor license agreements.  See the NOTICE file
# distributed with this work for additional information
# regarding copyright ownership.  The ASF licenses this file
# to you under the Apache License, Version 2.0 (the
# "License"); you may not use this file except in compliance
# with the License.  You may obtain a copy of the License at
#
#   http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing,
# software distributed under the License is distributed on an
# "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY
# KIND, either express or implied.  See the License for the
# specific language governing permissions and limitations
# under the License.
"""Attention cells."""
from __future__ import absolute_import
from __future__ import print_function

__all__ = ['AttentionCell', 'MultiHeadAttentionCell', 'MLPAttentionCell', 'DotProductAttentionCell']

import math
import mxnet as mx
from mxnet.gluon.block import HybridBlock
from mxnet.gluon import nn
from .block import L2Normalization

# TODO(sxjscience) Add mask flag to softmax operator. Think about how to accelerate the kernel
def _masked_softmax(F, att_score, mask):
    """Ignore the masked elements when calculating the softmax

    Parameters
    ----------
    F : symbol or ndarray
    att_score : Symborl or NDArray
        Shape (batch_size, query_length, memory_length)
    mask : Symbol or NDArray or None
        Shape (batch_size, query_length, memory_length)
    Returns
    -------
    att_weights : Symborl or NDArray
        Shape (batch_size, query_length, memory_length)
    """
    if mask is not None:
        # Fill in the masked scores with a very small value
        att_score = F.where(mask, att_score, -1e18 * F.ones_like(att_score))
        att_weights = F.softmax(att_score, axis=-1) * mask
    else:
        att_weights = F.softmax(att_score, axis=-1)
    return att_weights


# TODO(sxjscience) In the future, we should support setting mask/att_weights as sparse tensors
[docs]class AttentionCell(HybridBlock): """Abstract class for attention cells. Extend the class to implement your own attention method. One typical usage is to define your own `_compute_weight()` function to calculate the weights:: cell = AttentionCell() out = cell(query, key, value, mask) """ def _compute_weight(self, F, query, key, mask=None): """Compute attention weights based on the query and the keys Parameters ---------- F : symbol or ndarray query : Symbol or NDArray The query vectors. Shape (batch_size, query_length, query_dim) key : Symbol or NDArray Key of the memory. Shape (batch_size, memory_length, key_dim) mask : Symbol or NDArray or None Mask the memory slots. Shape (batch_size, query_length, memory_length) Only contains 0 or 1 where 0 means that the memory slot will not be used. If set to None. No mask will be used. Returns ------- att_weights : Symbol or NDArray For single-head attention, Shape (batch_size, query_length, memory_length) For multi-head attentino, Shape (batch_size, num_heads, query_length, memory_length) """ raise NotImplementedError def _read_by_weight(self, F, att_weights, value): """Read from the value matrix given the attention weights. Parameters ---------- F : symbol or ndarray att_weights : Symbol or NDArray Attention weights. For single-head attention, Shape (batch_size, query_length, memory_length). For multi-head attention, Shape (batch_size, num_heads, query_length, memory_length). value : Symbol or NDArray Value of the memory. Shape (batch_size, memory_length, total_value_dim) Returns ------- context_vec: Symbol or NDArray Shape (batch_size, query_length, context_vec_dim) """ return F.batch_dot(att_weights, value) def __call__(self, query, key, value=None, mask=None): # pylint: disable=arguments-differ """Compute the attention. Parameters ---------- query : Symbol or NDArray Query vector. Shape (batch_size, query_length, query_dim) key : Symbol or NDArray Key of the memory. Shape (batch_size, memory_length, key_dim) value : Symbol or NDArray or None, default None Value of the memory. If set to None, the value will be set as the key. Shape (batch_size, memory_length, value_dim) mask : Symbol or NDArray or None, default None Mask of the memory slots. Shape (batch_size, query_length, memory_length) Only contains 0 or 1 where 0 means that the memory slot will not be used. If set to None. No mask will be used. Returns ------- context_vec : Symbol or NDArray Shape (batch_size, query_length, context_vec_dim) att_weights : Symbol or NDArray Attention weights. Shape (batch_size, query_length, memory_length) """ return super(AttentionCell, self).__call__(query, key, value, mask)
[docs] def forward(self, query, key, value=None, mask=None): # pylint: disable=arguments-differ if value is None: value = key if mask is None: return super(AttentionCell, self).forward(query, key, value) else: return super(AttentionCell, self).forward(query, key, value, mask)
[docs] def hybrid_forward(self, F, query, key, value, mask=None): # pylint: disable=arguments-differ att_weights = self._compute_weight(F, query, key, mask) context_vec = self._read_by_weight(F, att_weights, value) return context_vec, att_weights
[docs]class MultiHeadAttentionCell(AttentionCell): r"""Multi-head Attention Cell. In the MultiHeadAttentionCell, the input query/key/value will be linearly projected for `num_heads` times with different projection matrices. Each projected key, value, query will be used to calculate the attention weights and values. The output of each head will be concatenated to form the final output. The idea is first proposed in "[Arxiv2014] Neural Turing Machines" and is later adopted in "[NIPS2017] Attention is All You Need" to solve the Neural Machine Translation problem. Parameters ---------- base_cell : AttentionCell query_units : int Total number of projected units for query. Must be divided exactly by num_heads. key_units : int Total number of projected units for key. Must be divided exactly by num_heads. value_units : int Total number of projected units for value. Must be divided exactly by num_heads. num_heads : int Number of parallel attention heads use_bias : bool, default True Whether to use bias when projecting the query/key/values weight_initializer : str or `Initializer` or None, default None Initializer of the weights. bias_initializer : str or `Initializer`, default 'zeros' Initializer of the bias. prefix : str or None, default None See document of `Block`. params : str or None, default None See document of `Block`. """ def __init__(self, base_cell, query_units, key_units, value_units, num_heads, use_bias=True, weight_initializer=None, bias_initializer='zeros', prefix=None, params=None): super(MultiHeadAttentionCell, self).__init__(prefix=prefix, params=params) self._base_cell = base_cell self._query_units = query_units self._key_units = key_units self._value_units = value_units self._num_heads = num_heads self._use_bias = use_bias if self._query_units % self._num_heads != 0: raise ValueError('In MultiHeadAttetion, the query_units should be divided exactly' ' by the number of heads. Received query_units={}, num_heads={}' .format(key_units, num_heads)) if self._key_units % self._num_heads != 0: raise ValueError('In MultiHeadAttetion, the key_units should be divided exactly' ' by the number of heads. Received key_units={}, num_heads={}' .format(key_units, num_heads)) if self._value_units % self._num_heads != 0: raise ValueError('In MultiHeadAttetion, the value_units should be divided exactly' ' by the number of heads. Received value_units={}, num_heads={}' .format(value_units, num_heads)) with self.name_scope(): self.proj_query = nn.Dense(units=self._query_units, use_bias=self._use_bias, flatten=False, weight_initializer=weight_initializer, bias_initializer=bias_initializer, prefix='query_') self.proj_key = nn.Dense(units=self._key_units, use_bias=self._use_bias, flatten=False, weight_initializer=weight_initializer, bias_initializer=bias_initializer, prefix='key_') self.proj_value = nn.Dense(units=self._value_units, use_bias=self._use_bias, flatten=False, weight_initializer=weight_initializer, bias_initializer=bias_initializer, prefix='value_') def __call__(self, query, key, value=None, mask=None): """Compute the attention. Parameters ---------- query : Symbol or NDArray Query vector. Shape (batch_size, query_length, query_dim) key : Symbol or NDArray Key of the memory. Shape (batch_size, memory_length, key_dim) value : Symbol or NDArray or None, default None Value of the memory. If set to None, the value will be set as the key. Shape (batch_size, memory_length, value_dim) mask : Symbol or NDArray or None, default None Mask of the memory slots. Shape (batch_size, query_length, memory_length) Only contains 0 or 1 where 0 means that the memory slot will not be used. If set to None. No mask will be used. Returns ------- context_vec : Symbol or NDArray Shape (batch_size, query_length, context_vec_dim) att_weights : Symbol or NDArray Attention weights of multiple heads. Shape (batch_size, num_heads, query_length, memory_length) """ return super(MultiHeadAttentionCell, self).__call__(query, key, value, mask) def _compute_weight(self, F, query, key, mask=None): query = self.proj_query(query) # Shape (batch_size, query_length, query_units) # Shape (batch_size * num_heads, query_length, ele_units) query = F.transpose(query.reshape(shape=(0, 0, self._num_heads, -1)), axes=(0, 2, 1, 3))\ .reshape(shape=(-1, 0, 0), reverse=True) key = self.proj_key(key) key = F.transpose(key.reshape(shape=(0, 0, self._num_heads, -1)), axes=(0, 2, 1, 3)).reshape(shape=(-1, 0, 0), reverse=True) if mask is not None: mask = F.broadcast_axis(F.expand_dims(mask, axis=1), axis=1, size=self._num_heads)\ .reshape(shape=(-1, 0, 0), reverse=True) att_weights = self._base_cell._compute_weight(F, query, key, mask) return att_weights.reshape(shape=(-1, self._num_heads, 0, 0), reverse=True) def _read_by_weight(self, F, att_weights, value): att_weights = att_weights.reshape(shape=(-1, 0, 0), reverse=True) value = self.proj_value(value) value = F.transpose(value.reshape(shape=(0, 0, self._num_heads, -1)), axes=(0, 2, 1, 3)).reshape(shape=(-1, 0, 0), reverse=True) context_vec = self._base_cell._read_by_weight(F, att_weights, value) context_vec = F.transpose(context_vec.reshape(shape=(-1, self._num_heads, 0, 0), reverse=True), axes=(0, 2, 1, 3)).reshape(shape=(0, 0, -1)) return context_vec
[docs]class MLPAttentionCell(AttentionCell): r"""Concat the query and the key and use a single-hidden-layer MLP to get the attention score. We provide two mode, the standard mode and the normalized mode. In the standard mode:: score = v tanh(W [h_q, h_k] + b) In the normalized mode (Same as TensorFlow):: score = g v / ||v||_2 tanh(W [h_q, h_k] + b) This type of attention is first proposed in .. Bahdanau et al., Neural Machine Translation by Jointly Learning to Align and Translate. ICLR 2015 Parameters ---------- units : int act : Activation, default nn.Activation('tanh') normalized : bool, default False Whether to normalize the weight that maps the embedded hidden states to the final score. This strategy can be interpreted as a type of "[NIPS2016] Weight Normalization". dropout : float, default 0.0 Attention dropout. weight_initializer : str or `Initializer` or None, default None Initializer of the weights. bias_initializer : str or `Initializer`, default 'zeros' Initializer of the bias. prefix : str or None, default None See document of `Block`. params : ParameterDict or None, default None See document of `Block`. """ def __init__(self, units, act=nn.Activation('tanh'), normalized=False, dropout=0.0, weight_initializer=None, bias_initializer='zeros', prefix=None, params=None): # Define a temporary class to implement the normalized version # TODO(sxjscience) Find a better solution class _NormalizedScoreProj(HybridBlock): def __init__(self, in_units, weight_initializer=None, prefix=None, params=None): super(_NormalizedScoreProj, self).__init__(prefix=prefix, params=params) self.g = self.params.get('g', shape=(1,), init=mx.init.Constant(1.0 / math.sqrt(in_units)), allow_deferred_init=True) self.v = self.params.get('v', shape=(1, in_units), init=weight_initializer, allow_deferred_init=True) def hybrid_forward(self, F, x, g, v): # pylint: disable=arguments-differ v = F.broadcast_div(v, F.sqrt(F.dot(v, v, transpose_b=True))) weight = F.broadcast_mul(g, v) out = F.FullyConnected(x, weight, None, no_bias=True, num_hidden=1, flatten=False, name='fwd') return out super(MLPAttentionCell, self).__init__(prefix=prefix, params=params) self._units = units self._act = act self._normalized = normalized self._dropout = dropout with self.name_scope(): self._dropout_layer = nn.Dropout(dropout) self._query_mid_layer = nn.Dense(units=self._units, flatten=False, use_bias=True, weight_initializer=weight_initializer, bias_initializer=bias_initializer, prefix='query_') self._key_mid_layer = nn.Dense(units=self._units, flatten=False, use_bias=False, weight_initializer=weight_initializer, prefix='key_') if self._normalized: self._attention_score = \ _NormalizedScoreProj(in_units=units, weight_initializer=weight_initializer, prefix='score_') else: self._attention_score = nn.Dense(units=1, in_units=self._units, flatten=False, use_bias=False, weight_initializer=weight_initializer, prefix='score_') def _compute_weight(self, F, query, key, mask=None): mapped_query = self._query_mid_layer(query) mapped_key = self._key_mid_layer(key) mid_feat = F.broadcast_add(F.expand_dims(mapped_query, axis=2), F.expand_dims(mapped_key, axis=1)) mid_feat = self._act(mid_feat) att_score = self._attention_score(mid_feat).reshape(shape=(0, 0, 0)) att_weights = self._dropout_layer(_masked_softmax(F, att_score, mask)) return att_weights
[docs]class DotProductAttentionCell(AttentionCell): r"""Dot product attention between the query and the key. Depending on parameters, defined as:: units is None: score = <h_q, h_k> units is not None and luong_style is False: score = <W_q h_q, W_k h_k> units is not None and luong_style is True: score = <W h_q, h_k> Parameters ---------- units: int or None, default None Project the query and key to vectors with `units` dimension before applying the attention. If set to None, the query vector and the key vector are directly used to compute the attention and should have the same dimension:: If the units is None, score = <h_q, h_k> Else if the units is not None and luong_style is False: score = <W_q h_q, W_k, h_k> Else if the units is not None and luong_style is True: score = <W h_q, h_k> luong_style: bool, default False If turned on, the score will be:: score = <W h_q, h_k> `units` must be the same as the dimension of the key vector scaled: bool, default True Whether to divide the attention weights by the sqrt of the query dimension. This is first proposed in "[NIPS2017] Attention is all you need.":: score = <h_q, h_k> / sqrt(dim_q) normalized: bool, default False If turned on, the cosine distance is used, i.e:: score = <h_q / ||h_q||, h_k / ||h_k||> use_bias : bool, default True Whether to use bias in the projection layers. dropout : float, default 0.0 Attention dropout weight_initializer : str or `Initializer` or None, default None Initializer of the weights bias_initializer : str or `Initializer`, default 'zeros' Initializer of the bias prefix : str or None, default None See document of `Block`. params : str or None, default None See document of `Block`. """ def __init__(self, units=None, luong_style=False, scaled=True, normalized=False, use_bias=True, dropout=0.0, weight_initializer=None, bias_initializer='zeros', prefix=None, params=None): super(DotProductAttentionCell, self).__init__(prefix=prefix, params=params) self._units = units self._scaled = scaled self._normalized = normalized self._use_bias = use_bias self._luong_style = luong_style self._dropout = dropout if self._luong_style: assert units is not None, 'Luong style attention is not available without explicitly ' \ 'setting the units' with self.name_scope(): self._dropout_layer = nn.Dropout(dropout) if units is not None: with self.name_scope(): self._proj_query = nn.Dense(units=self._units, use_bias=self._use_bias, flatten=False, weight_initializer=weight_initializer, bias_initializer=bias_initializer, prefix='query_') if not self._luong_style: self._proj_key = nn.Dense(units=self._units, use_bias=self._use_bias, flatten=False, weight_initializer=weight_initializer, bias_initializer=bias_initializer, prefix='key_') if self._normalized: with self.name_scope(): self._l2_norm = L2Normalization(axis=-1) def _compute_weight(self, F, query, key, mask=None): if self._units is not None: query = self._proj_query(query) if not self._luong_style: key = self._proj_key(key) elif F == mx.nd: assert query.shape[-1] == key.shape[-1], 'Luong style attention requires key to ' \ 'have the same dim as the projected ' \ 'query. Received key {}, query {}.'.format( key.shape, query.shape) if self._normalized: query = self._l2_norm(query) key = self._l2_norm(key) if self._scaled: query = F.contrib.div_sqrt_dim(query) att_score = F.batch_dot(query, key, transpose_b=True) att_weights = self._dropout_layer(_masked_softmax(F, att_score, mask)) return att_weights