260 lines
10 KiB
Python
260 lines
10 KiB
Python
|
# Copyright 2019 The TensorFlow Authors. All Rights Reserved.
|
||
|
|
#
|
||
|
|
# Licensed 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.
|
||
|
|
# ==============================================================================
|
||
|
|
"""Base class for attention layers that can be used in sequence DNN/CNN models.
|
||
|
|
|
||
|
|
This file follows the terminology of https://arxiv.org/abs/1706.03762 Figure 2.
|
||
|
|
Attention is formed by three tensors: Query, Key and Value.
|
||
|
|
"""
|
||
|
|
|
||
|
|
import tensorflow.compat.v2 as tf
|
||
|
|
from absl import logging
|
||
|
|
|
||
|
|
from keras import backend
|
||
|
|
from keras.engine import base_layer
|
||
|
|
from keras.utils import control_flow_util
|
||
|
|
|
||
|
|
# isort: off
|
||
|
|
from tensorflow.python.util.tf_export import keras_export
|
||
|
|
|
||
|
|
|
||
|
|
@keras_export("keras.__internal__.layers.BaseDenseAttention", v1=[])
|
||
|
|
class BaseDenseAttention(base_layer.BaseRandomLayer):
|
||
|
|
"""Base Attention class for Dense networks.
|
||
|
|
|
||
|
|
This class is suitable for Dense or CNN networks, and not for RNN networks.
|
||
|
|
|
||
|
|
Implementations of attention mechanisms should inherit from this class, and
|
||
|
|
reuse the `apply_attention_scores()` method.
|
||
|
|
|
||
|
|
Args:
|
||
|
|
dropout: Float between 0 and 1. Fraction of the units to drop for the
|
||
|
|
attention scores.
|
||
|
|
|
||
|
|
Call Args:
|
||
|
|
inputs: List of the following tensors:
|
||
|
|
* query: Query `Tensor` of shape `[batch_size, Tq, dim]`.
|
||
|
|
* value: Value `Tensor` of shape `[batch_size, Tv, dim]`.
|
||
|
|
* key: Optional key `Tensor` of shape `[batch_size, Tv, dim]`. If not
|
||
|
|
given, will use `value` for both `key` and `value`, which is the most
|
||
|
|
common case.
|
||
|
|
mask: List of the following tensors:
|
||
|
|
* query_mask: A boolean mask `Tensor` of shape `[batch_size, Tq]`. If
|
||
|
|
given, the output will be zero at the positions where `mask==False`.
|
||
|
|
* value_mask: A boolean mask `Tensor` of shape `[batch_size, Tv]`. If
|
||
|
|
given, will apply the mask such that values at positions where
|
||
|
|
`mask==False` do not contribute to the result.
|
||
|
|
training: Python boolean indicating whether the layer should behave in
|
||
|
|
training mode (adding dropout) or in inference mode (no dropout).
|
||
|
|
return_attention_scores: bool, if `True`, returns the attention scores
|
||
|
|
(after masking and softmax) as an additional output argument.
|
||
|
|
|
||
|
|
Output:
|
||
|
|
|
||
|
|
Attention outputs of shape `[batch_size, Tq, dim]`.
|
||
|
|
[Optional] Attention scores after masking and softmax with shape
|
||
|
|
`[batch_size, Tq, Tv]`.
|
||
|
|
"""
|
||
|
|
|
||
|
|
def __init__(self, dropout=0.0, **kwargs):
|
||
|
|
# Deprecated field `causal` determines whether to using causal masking.
|
||
|
|
# Use `use_causal_mask` in call() method instead.
|
||
|
|
if "causal" in kwargs:
|
||
|
|
logging.warning(
|
||
|
|
"`causal` argument is deprecated. Please use `use_causal_mask` "
|
||
|
|
"in call() method to specify causal masking."
|
||
|
|
)
|
||
|
|
self.causal = kwargs.pop("causal", False)
|
||
|
|
super().__init__(**kwargs)
|
||
|
|
self.dropout = dropout
|
||
|
|
self.supports_masking = True
|
||
|
|
|
||
|
|
def build(self, input_shape):
|
||
|
|
# Skip RNG initialization if dropout rate is 0. This will let the layer
|
||
|
|
# be purely stateless, with no reference to any variable.
|
||
|
|
if self.dropout > 0:
|
||
|
|
super().build(input_shape)
|
||
|
|
self.built = True
|
||
|
|
|
||
|
|
def _calculate_scores(self, query, key):
|
||
|
|
"""Calculates attention scores.
|
||
|
|
|
||
|
|
Args:
|
||
|
|
query: Query tensor of shape `[batch_size, Tq, dim]`.
|
||
|
|
key: Key tensor of shape `[batch_size, Tv, dim]`.
|
||
|
|
|
||
|
|
Returns:
|
||
|
|
Tensor of shape `[batch_size, Tq, Tv]`.
|
||
|
|
"""
|
||
|
|
return NotImplementedError
|
||
|
|
|
||
|
|
def _apply_scores(self, scores, value, scores_mask=None, training=None):
|
||
|
|
"""Applies attention scores to the given value tensor.
|
||
|
|
|
||
|
|
To use this method in your attention layer, follow the steps:
|
||
|
|
|
||
|
|
* Use `query` tensor of shape `[batch_size, Tq]` and `key` tensor of
|
||
|
|
shape `[batch_size, Tv]` to calculate the attention `scores`.
|
||
|
|
* Pass `scores` and `value` tensors to this method. The method applies
|
||
|
|
`scores_mask`, calculates `attention_distribution = softmax(scores)`,
|
||
|
|
then returns `matmul(attention_distribution, value).
|
||
|
|
* Apply `query_mask` and return the result.
|
||
|
|
|
||
|
|
Args:
|
||
|
|
scores: Scores float tensor of shape `[batch_size, Tq, Tv]`.
|
||
|
|
value: Value tensor of shape `[batch_size, Tv, dim]`.
|
||
|
|
scores_mask: A boolean mask `Tensor` of shape `[batch_size, 1, Tv]` or
|
||
|
|
`[batch_size, Tq, Tv]`. If given, scores at positions where
|
||
|
|
`scores_mask==False` do not contribute to the result. It must
|
||
|
|
contain at least one `True` value in each line along the last
|
||
|
|
dimension.
|
||
|
|
training: Python boolean indicating whether the layer should behave in
|
||
|
|
training mode (adding dropout) or in inference mode (no dropout).
|
||
|
|
|
||
|
|
Returns:
|
||
|
|
Tensor of shape `[batch_size, Tq, dim]`.
|
||
|
|
Attention scores after masking and softmax with shape
|
||
|
|
`[batch_size, Tq, Tv]`.
|
||
|
|
"""
|
||
|
|
if scores_mask is not None:
|
||
|
|
padding_mask = tf.logical_not(scores_mask)
|
||
|
|
# Bias so padding positions do not contribute to attention
|
||
|
|
# distribution. Note 65504. is the max float16 value.
|
||
|
|
if scores.dtype is tf.float16:
|
||
|
|
scores -= 65504.0 * tf.cast(padding_mask, dtype=scores.dtype)
|
||
|
|
else:
|
||
|
|
scores -= 1.0e9 * tf.cast(padding_mask, dtype=scores.dtype)
|
||
|
|
if training is None:
|
||
|
|
training = backend.learning_phase()
|
||
|
|
weights = tf.nn.softmax(scores)
|
||
|
|
|
||
|
|
if self.dropout > 0:
|
||
|
|
|
||
|
|
def dropped_weights():
|
||
|
|
return self._random_generator.dropout(
|
||
|
|
weights, rate=self.dropout
|
||
|
|
)
|
||
|
|
|
||
|
|
weights = control_flow_util.smart_cond(
|
||
|
|
training, dropped_weights, lambda: tf.identity(weights)
|
||
|
|
)
|
||
|
|
return tf.matmul(weights, value), weights
|
||
|
|
|
||
|
|
# TODO(b/125916026): Consider exposing a __call__ method with named args.
|
||
|
|
def call(
|
||
|
|
self,
|
||
|
|
inputs,
|
||
|
|
mask=None,
|
||
|
|
training=None,
|
||
|
|
return_attention_scores=False,
|
||
|
|
use_causal_mask=False,
|
||
|
|
):
|
||
|
|
self._validate_call_args(inputs=inputs, mask=mask)
|
||
|
|
q = inputs[0]
|
||
|
|
v = inputs[1]
|
||
|
|
k = inputs[2] if len(inputs) > 2 else v
|
||
|
|
q_mask = mask[0] if mask else None
|
||
|
|
v_mask = mask[1] if mask else None
|
||
|
|
scores = self._calculate_scores(query=q, key=k)
|
||
|
|
if v_mask is not None:
|
||
|
|
# Mask of shape [batch_size, 1, Tv].
|
||
|
|
v_mask = tf.expand_dims(v_mask, axis=-2)
|
||
|
|
if self.causal or use_causal_mask:
|
||
|
|
# Creates a lower triangular mask, so position i cannot attend to
|
||
|
|
# positions j>i. This prevents the flow of information from the
|
||
|
|
# future into the past.
|
||
|
|
scores_shape = tf.shape(scores)
|
||
|
|
# causal_mask_shape = [1, Tq, Tv].
|
||
|
|
causal_mask_shape = tf.concat(
|
||
|
|
[tf.ones_like(scores_shape[:-2]), scores_shape[-2:]], axis=0
|
||
|
|
)
|
||
|
|
causal_mask = _lower_triangular_mask(causal_mask_shape)
|
||
|
|
else:
|
||
|
|
causal_mask = None
|
||
|
|
scores_mask = _merge_masks(v_mask, causal_mask)
|
||
|
|
result, attention_scores = self._apply_scores(
|
||
|
|
scores=scores, value=v, scores_mask=scores_mask, training=training
|
||
|
|
)
|
||
|
|
if q_mask is not None:
|
||
|
|
# Mask of shape [batch_size, Tq, 1].
|
||
|
|
q_mask = tf.expand_dims(q_mask, axis=-1)
|
||
|
|
result *= tf.cast(q_mask, dtype=result.dtype)
|
||
|
|
if return_attention_scores:
|
||
|
|
return result, attention_scores
|
||
|
|
return result
|
||
|
|
|
||
|
|
def compute_mask(self, inputs, mask=None):
|
||
|
|
self._validate_call_args(inputs=inputs, mask=mask)
|
||
|
|
if mask:
|
||
|
|
q_mask = mask[0]
|
||
|
|
if q_mask is None:
|
||
|
|
return None
|
||
|
|
return tf.convert_to_tensor(q_mask)
|
||
|
|
return None
|
||
|
|
|
||
|
|
def compute_output_shape(self, input_shape):
|
||
|
|
# return_attention_scores argument of BaseDenseAttention.call method
|
||
|
|
# is ignored. Output shape of attention_scores cannot be returned.
|
||
|
|
return tf.TensorShape(input_shape[0])
|
||
|
|
|
||
|
|
def _validate_call_args(self, inputs, mask):
|
||
|
|
"""Validates arguments of the call method."""
|
||
|
|
class_name = self.__class__.__name__
|
||
|
|
if not isinstance(inputs, list):
|
||
|
|
raise ValueError(
|
||
|
|
f"{class_name} layer must be called on a list of inputs, "
|
||
|
|
"namely [query, value] or [query, value, key]. "
|
||
|
|
f"Received: {inputs}."
|
||
|
|
)
|
||
|
|
if len(inputs) < 2 or len(inputs) > 3:
|
||
|
|
raise ValueError(
|
||
|
|
f"{class_name} layer accepts inputs list of length 2 or 3, "
|
||
|
|
"namely [query, value] or [query, value, key]. "
|
||
|
|
f"Received length: {len(inputs)}."
|
||
|
|
)
|
||
|
|
if mask:
|
||
|
|
if not isinstance(mask, list):
|
||
|
|
raise ValueError(
|
||
|
|
f"{class_name} layer mask must be a list, "
|
||
|
|
f"namely [query_mask, value_mask]. Received: {mask}."
|
||
|
|
)
|
||
|
|
if len(mask) < 2 or len(mask) > len(inputs):
|
||
|
|
raise ValueError(
|
||
|
|
f"{class_name} layer mask must be a list of length 2, "
|
||
|
|
"namely [query_mask, value_mask]. "
|
||
|
|
f"Received length: {len(mask)}."
|
||
|
|
)
|
||
|
|
|
||
|
|
def get_config(self):
|
||
|
|
config = {
|
||
|
|
"dropout": self.dropout,
|
||
|
|
}
|
||
|
|
base_config = super().get_config()
|
||
|
|
return dict(list(base_config.items()) + list(config.items()))
|
||
|
|
|
||
|
|
|
||
|
|
def _lower_triangular_mask(shape):
|
||
|
|
"""Creates a lower-triangular boolean mask over the last 2 dimensions."""
|
||
|
|
row_index = tf.cumsum(tf.ones(shape=shape, dtype=tf.int32), axis=-2)
|
||
|
|
col_index = tf.cumsum(tf.ones(shape=shape, dtype=tf.int32), axis=-1)
|
||
|
|
return tf.greater_equal(row_index, col_index)
|
||
|
|
|
||
|
|
|
||
|
|
def _merge_masks(x, y):
|
||
|
|
if x is None:
|
||
|
|
return y
|
||
|
|
if y is None:
|
||
|
|
return x
|
||
|
|
return tf.logical_and(x, y)
|