File size: 3,355 Bytes
3ce19a2 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 | """
-*- coding: utf-8 -*-
File : batchnorm_reimpl.py
Author : Jiayuan Mao
Email : maojiayuan@gmail.com
Date : 27/01/2018
This file is part of Synchronized-BatchNorm-PyTorch.
https://github.com/vacancy/Synchronized-BatchNorm-PyTorch
Distributed under MIT License.
MIT License
Copyright (c) 2018 Jiayuan MAO
Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software"), to deal
in the Software without restriction, including without limitation the rights
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all
copies or substantial portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
SOFTWARE.
"""
import torch
import torch.nn as nn
import torch.nn.init as init
__all__ = ['BatchNorm2dReimpl']
class BatchNorm2dReimpl(nn.Module):
"""
A re-implementation of batch normalization, used for testing the numerical
stability.
Author: acgtyrant
See also:
https://github.com/vacancy/Synchronized-BatchNorm-PyTorch/issues/14
"""
def __init__(self, num_features, eps=1e-5, momentum=0.1):
super().__init__()
self.num_features = num_features
self.eps = eps
self.momentum = momentum
self.weight = nn.Parameter(torch.empty(num_features))
self.bias = nn.Parameter(torch.empty(num_features))
self.register_buffer('running_mean', torch.zeros(num_features))
self.register_buffer('running_var', torch.ones(num_features))
self.reset_parameters()
def reset_running_stats(self):
self.running_mean.zero_()
self.running_var.fill_(1)
def reset_parameters(self):
self.reset_running_stats()
init.uniform_(self.weight)
init.zeros_(self.bias)
def forward(self, input_):
batchsize, channels, height, width = input_.size()
numel = batchsize * height * width
input_ = input_.permute(1, 0, 2, 3).contiguous().view(channels, numel)
sum_ = input_.sum(1)
sum_of_square = input_.pow(2).sum(1)
mean = sum_ / numel
sumvar = sum_of_square - sum_ * mean
self.running_mean = ((1 - self.momentum) * self.running_mean + self.momentum * mean.detach())
unbias_var = sumvar / (numel - 1)
self.running_var = ((1 - self.momentum) * self.running_var + self.momentum * unbias_var.detach())
bias_var = sumvar / numel
inv_std = 1 / (bias_var + self.eps).pow(0.5)
output = ((input_ - mean.unsqueeze(1)) * inv_std.unsqueeze(1) * self.weight.unsqueeze(1) +
self.bias.unsqueeze(1))
return output.view(channels, batchsize, height, width).permute(1, 0, 2, 3).contiguous()
|