# Copyright (c) Microsoft Corporation. All rights reserved.
# Licensed under the MIT license.
import torch
import torch.nn as nn
import logging
from ..BaseLayer import BaseConf
from utils.DocInherit import DocInherit
from utils.exceptions import ConfigurationError
import copy
[docs]class Minus2DConf(BaseConf):
""" Configuration of Minus2D layer
Args:
abs_flag: if the result of the Minus2D is abs, default is False
"""
#init the args
def __init__(self,**kwargs):
super(Minus2DConf, self).__init__(**kwargs)
#set default params
[docs] @DocInherit
def default(self):
self.abs_flag = False
[docs] @DocInherit
def declare(self):
self.num_of_inputs = 2
self.input_ranks = [2,2]
[docs] @DocInherit
def inference(self):
self.output_dim = copy.deepcopy(self.input_dims[0])
if self.input_dims[0][1] != 1:
self.output_dim[-1] = self.input_dims[0][1]
else:
self.output_dim[-1] = self.input_dims[1][1]
super(Minus2DConf, self).inference()
[docs] @DocInherit
def verify(self):
super(Minus2DConf, self).verify()
# # to check if the ranks of all the inputs are equal
# rank_equal_flag = True
# for i in range(len(self.input_ranks)):
# if self.input_ranks[i] != self.input_ranks[0]:
# rank_equal_flag = False
# break
# if rank_equal_flag == False:
# raise ConfigurationError("For layer Minus2D, the ranks of each inputs should be equal!")
# to check if the dimensions of all the inputs are equal or is 1
dim_flag = True
input_dims = list(self.input_dims)
for i in range(len(input_dims)):
if input_dims[i][1] != input_dims[0][1] and input_dims[i][1] != 1 and input_dims[0][1] != 1:
dim_flag = False
break
if dim_flag == False:
raise ConfigurationError("For layer Minus2D, the dimensions of each inputs should be equal or 1")
[docs]class Minus2D(nn.Module):
"""Minus2D layer to get subtraction of two sequences(2D representation)
Args:
layer_conf (Minus2DConf): configuration of a layer
"""
def __init__(self, layer_conf):
super(Minus2D, self).__init__()
self.layer_conf = layer_conf
logging.warning("The length Minus2D layer returns is the length of first input")
[docs] def forward(self, *args):
""" process inputs
Args:
*args: (Tensor): string, string_len, string2, string2_len
e.g. string (Tensor): [batch_size, dim], string_len (Tensor): [batch_size]
Returns:
Tensor: [batch_size, output_dim], [batch_size]
"""
if self.layer_conf.abs_flag == False:
return (args[0] - args[2]), args[1]
if self.layer_conf.abs_flag == True:
return torch.abs(args[0] - args[2]),args[1]