ICode9

精准搜索请尝试: 精确搜索
首页 > 其他分享> 文章详细

Pytorch 损失函数总结

2022-03-09 09:32:58  阅读:318  来源: 互联网

标签:总结 loss 函数 nn torch reduction Pytorch output tensor


1 nn.L1Loss

  torch.nn.L1Loss(reduction='mean')

  就是 MAE(mean absolute error),计算公式为

    $\ell(x, y)=L=\left\{l_{1}, \ldots, l_{N}\right\}^{\top}, \quad l_{n}=\left|x_{n}-y_{n}\right|$

    $\ell(x, y)=\left\{\begin{array}{ll}\operatorname{mean}(L), & \text { if reduction }=\text { 'mean'; } \\\operatorname{sum}(L), & \text { if reduction }=\text { 'sum' }\end{array}\right.$

  例子:逐元素计算

input = torch.arange(1,7.).view(2,3)
target = torch.arange(6).view(2,3)
print(input)
print(target)
"""
tensor([[1., 2., 3.],
        [4., 5., 6.]])
tensor([[0, 1, 2],
        [3, 4, 5]])
"""
loss = nn.L1Loss(reduction='sum')
output = loss(input, target)
print(output)
"""
tensor(6.)
"""
loss = nn.L1Loss(reduction='mean')
output = loss(input, target)
print(output)
"""
tensor(1.)
"""

2 nn.MSELoss

    torch.nn.MSELoss(reduction='mean')

  如其名,mean squared error,也就是 L2 正则项,计算公式为

  $\ell(x, y)=\left\{\begin{array}{ll}\operatorname{mean}(L), & \text { if reduction }=\text { 'mean'; } \\\operatorname{sum}(L), & \text { if reduction }=\text { 'sum' }\end{array}\right.$

  $\ell(x, y)=L=\left\{l_{1}, \ldots, l_{N}\right\}^{\top}, \quad l_{n}=\left(x_{n}-y_{n}\right)^{2}$

  有 mean 和 sum 两种模式选,通过 reduction 控制。

  例子:逐元素计算

loss = nn.MSELoss(reduction="mean")
output = loss(input, target)
print(output)
"""
tensor(1.)
"""
loss = nn.MSELoss(reduction="sum")
output = loss(input, target)
print(output)
"""
tensor(6.)
"""

  从上述实验可以看出

    $l_{n}=\left(x_{n}-y_{n}\right)^{2}$ 

  是逐元素计算。

3 nn.SmoothL1Loss

    torch.nn.SmoothL1Loss(reduction='mean', beta=1.0)

  对 L1 做了一点平滑,比起MSELoss,对于 outlier 更加不敏感。

    $\ell(x, y)=L=\left\{l_{1}, \ldots, l_{N}\right\}^{T}$

    $l_{n}=\left\{\begin{array}{ll}0.5\left(x_{n}-y_{n}\right)^{2} / \text { beta }, & \text { if }\left|x_{n}-y_{n}\right|<\text { beta } \\\left|x_{n}-y_{n}\right|-0.5 * \text { beta }, & \text { otherwise }\end{array}\right.$

  在Fast-RCNN中使用以避免梯度爆炸。

  例子:逐元素计算

loss = nn.MSELoss(reduction="sum")
output = loss(input, target)
print(output)
"""
tensor(6.)
"""
loss = nn.SmoothL1Loss(reduction="mean")
output = loss(input, target)
print(output)
"""
tensor(0.5000)
"""
loss = nn.SmoothL1Loss(reduction="mean",beta = 3)
output = loss(input, target)
print(output)
"""
tensor(0.1667)
"""

4 nn.BCELoss 以及 nn.BCEWithLogitsLoss

    torch.nn.BCELoss(weight=None,reduction='mean')

  Binary Cross Entropy,公式如下:

    $\ell(x, y)=\left\{\begin{array}{ll}\operatorname{mean}(L), & \text { if reduction }=\text { 'mean'; } \\\operatorname{sum}(L), & \text { if reduction }=\text { 'sum' }\end{array}\right.$

    $\ell(x, y)=L=\left\{l_{1}, \ldots, l_{N}\right\}^{\top}, \quad l_{n}=-w_{n}\left[y_{n} \cdot \log x_{n}+\left(1-y_{n}\right) \cdot \log \left(1-x_{n}\right)\right]$

  双向的交叉熵,相当于交叉熵公式的二分类简化版,可以用于分类不互斥的多分类任务。

  BCELoss需要先手动对输入 sigmoid,然后每一个位置如果分类是 1 则加 $-log(exp(x))$ 否则加 $-log(exp(1-x))$,最后求取平均。

  BCEWithLogitsLoss 则不需要 sigmoid,其他都完全一样。

  例子:逐元素计算。

target = torch.tensor([[1,0,1],[0,1,1]],dtype = torch.float32)
raw_output = torch.randn(2,3,dtype = torch.float32)
output = torch.sigmoid(raw_output)
print(output)

result = np.zeros((2,3))
for ix in range(2):
    for iy in range(3):
        if(target[ix, iy]==1): 
            result[ix, iy] += -np.log(output[ix, iy])
        elif(target[ix, iy]==0): 
            result[ix, iy] += -np.log(1-output[ix, iy])

print(result)
print(np.mean(result))

loss_fn = torch.nn.BCELoss(reduction='none')
print(loss_fn(output, target))
loss_fn = torch.nn.BCELoss(reduction='mean')
print(loss_fn(output, target))
loss_fn = torch.nn.BCEWithLogitsLoss(reduction='sum')
print(loss_fn(raw_output, target))
tensor([[0.5316, 0.6816, 0.4768],
        [0.6485, 0.3037, 0.5490]])

[[0.63186073 1.14431179 0.74067789]
 [1.04543173 1.19187558 0.59973639]]

0.892315685749054

tensor([[0.6319, 1.1443, 0.7407],
        [1.0454, 1.1919, 0.5997]])

tensor(0.8923)

tensor(5.3539)

5 nn.CrossEntropyLoss

     torch.nn.CrossEntropyLoss(weight=None, ignore_index=- 100, reduction='mean', label_smoothing=0.0)

  经典Loss, 计算公式为:

    $\text { weight }[\text { class }]\left(-\log \left(\frac{\exp (x[\text { class }])}{\sum\limits_{j} \exp (x[j])}\right)\right)=\text { weight }[\text { class }]\left(-x[\text { class }]+\log \left(\sum\limits_{j} \exp (x[j])\right)\right)$

  相当于先将输出值通过 softmax 映射到每个值在 $[0,1]$,和为 $1$ 的空间上。

  希望正确的 class 对应的 loss 越小越好,所以对 $\left(\frac {\exp (x[\text {class}])}{\sum\limits _{j} \exp (x[j])}\right)$ 求取 $-log()$, 把 $[0,1]$ 映射到 $[0,+\infty]$ 上,正确项的概率占比越大,整体损失就越小。

  torch里的CrossEntropyLoss(x) 等价于 NLLLoss(LogSoftmax(x))

  预期输入未normalize过的score,输入形状和NLL一样,为$(N,C)和(N)$

  例子:按样本数计算

target = torch.tensor([1,0,3])
output = torch.randn(3,5)
print(output)
"""
tensor([[-2.5728, -0.4581, -0.2017,  1.8813,  0.4544],
        [-0.7278,  0.6300,  0.6510, -1.7570,  1.1788],
        [-0.4660,  0.0410,  0.6876,  0.8966,  0.1446]])
"""
loss_fn = torch.nn.CrossEntropyLoss(reduction='mean')
loss = loss_fn(output, target)
print(loss)
"""
tensor(2.1940)
"""
loss_fn = torch.nn.CrossEntropyLoss(reduction='sum')
loss = loss_fn(output, target)
print(loss)
"""
tensor(6.5821)
"""

   例子:手写版

target = torch.tensor([1,0,3])
output = torch.randn(3,5)
print(output)
"""
tensor([[-0.1168,  1.5417,  1.1748, -1.1856, -0.1233],
        [ 0.2074, -0.7376, -0.8934,  0.0899,  0.5337],
        [-0.5323, -0.2945, -0.1710,  1.5925,  1.3654]])
"""
result = np.array([0.0, 0.0, 0.0])
for ix in range(3):
    log_sum = 0.0
    for iy in range(5):
        if(iy==target[ix]): 
            result[ix] += -output[ix, iy]
        log_sum += np.exp(output[ix, iy])
    result[ix] += np.log(log_sum)
print(result)
print(np.mean(result))

loss_fn = torch.nn.CrossEntropyLoss(reduction='mean')
loss = loss_fn(output, target)
print(loss.item())
"""
[0.75984335 1.3853296  0.80614853]
0.9837738275527954
0.9837737679481506
"""

6 nn.NLLLoss

     torch.nn.NLLLoss(weight=None,ignore_index=- 100, reduction='mean')

  negative log likelihood loss,用于训练 n 类分类器,对于不平衡数据集,可以给类别添加 weight,计算公式为
    $l_{n}=-w_{y_{n}} x_{n, y_{n}}$

    $-w_{c}=\text { weight }[c] \cdot 1$

  预期输入形状 $(N,C)$ 和 $(N)$,其中 $N$ 为 batch 大小,$C$ 为类别数;

  计算每个 case 的 target 对应类别的概率的负值,然后求取平均/和,一般与一个 LogSoftMax 连用从而获得对数概率。

  例子:按样本数计算

target = torch.tensor([1,0,3])
output = torch.randn(3,5)
print(output)

loss_fn = torch.nn.NLLLoss(reduction='mean')
loss = loss_fn(output, target)
print(loss)

loss_fn = torch.nn.NLLLoss(reduction='sum')
loss = loss_fn(output, target)
print(loss)
"""
tensor([[ 1.5083,  0.1846, -1.8400, -0.0068, -0.1943],
        [ 0.5303, -0.0350, -0.3924,  0.3026,  0.6159],
        [ 2.0047, -1.0653,  0.0718, -0.8632, -1.0695]])
tensor(0.0494)
tensor(0.1482)
"""

  显然不是逐元素计算。

  例子:

import torch
input=torch.randn(3,3)
soft_input = torch.nn.Softmax(dim=0)
soft_input(input)
"""
tensor([[0.2603, 0.6519, 0.5811],
        [0.5248, 0.3026, 0.1783],
        [0.2148, 0.0455, 0.2406]])
"""
#对softmax结果取log
torch.log(soft_input(input))
"""
tensor([[-1.3458, -0.4279, -0.5428],
        [-0.6447, -1.1952, -1.7243],
        [-1.5379, -3.0898, -1.4248]])
"""

  假设标签是[0,1,2],第一行取第0个元素,第二行取第1个,第三行取第2个,去掉负号,即[0.3168,3.3093,0.4701],求平均值,就可以得到损失值。

(0.3168+3.3093+0.4701)/3
"""
1.3654000000000002
"""
loss=torch.nn.NLLLoss()
target=torch.tensor([0,1,2])
loss(input,target)
"""
tensor(-0.1395)
"""

   所以 nn.NLLLoss 计算方式为:log(softmax) 取平均

 

参考:https://segmentfault.com/a/1190000038584083

标签:总结,loss,函数,nn,torch,reduction,Pytorch,output,tensor
来源: https://www.cnblogs.com/BlairGrowing/p/15982528.html

本站声明: 1. iCode9 技术分享网(下文简称本站)提供的所有内容,仅供技术学习、探讨和分享;
2. 关于本站的所有留言、评论、转载及引用,纯属内容发起人的个人观点,与本站观点和立场无关;
3. 关于本站的所有言论和文字,纯属内容发起人的个人观点,与本站观点和立场无关;
4. 本站文章均是网友提供,不完全保证技术分享内容的完整性、准确性、时效性、风险性和版权归属;如您发现该文章侵犯了您的权益,可联系我们第一时间进行删除;
5. 本站为非盈利性的个人网站,所有内容不会用来进行牟利,也不会利用任何形式的广告来间接获益,纯粹是为了广大技术爱好者提供技术内容和技术思想的分享性交流网站。

专注分享技术,共同学习,共同进步。侵权联系[81616952@qq.com]

Copyright (C)ICode9.com, All Rights Reserved.

ICode9版权所有