可重现的 Pytorch 结果和随机种子

Question

我有一个带有 Pytorch 的简单玩具 NN。我正在设置我可以在文档中找到的所有种子以及随机的 numpy。

如果我运行下面的代码从上到下，结果似乎是可重现的。

但是，如果我运行块 1 仅一次，然后每次运行块 2，结果会发生变化（有时会发生显着变化）。我不确定为什么会发生这种情况，因为每次都会重新初始化网络和重置优化器。

我使用的是 0.4.0 版本

区块 #1

from __future__ import division
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd

import torch
import torch.utils.data as utils_data
from torch.autograd import Variable
from torch import optim, nn
from torch.utils.data import Dataset 
import torch.nn.functional as F
from torch.nn.init import xavier_uniform_, xavier_normal_,uniform_

torch.manual_seed(123)

import random
random.seed(123)


from sklearn.datasets import load_boston
from sklearn.metrics import mean_squared_error
from sklearn.model_selection import train_test_split

%matplotlib inline 

cuda=True #set to true uses GPU

if cuda:
    torch.cuda.manual_seed(123)

#load boston data from scikit
boston = load_boston()
x=boston.data
y=boston.target
y=y.reshape(y.shape[0],1)

#train and test
x_train, x_test, y_train, y_test = train_test_split(x,y,test_size=0.3, random_state=123, shuffle=False)


#change to tensors
x_train = torch.from_numpy(x_train)
y_train = torch.from_numpy(y_train)

#create dataset and use data loader
training_samples = utils_data.TensorDataset(x_train, y_train)
data_loader_trn = utils_data.DataLoader(training_samples, batch_size=64,drop_last=False)

#change to tensors
x_test = torch.from_numpy(x_test)
y_test = torch.from_numpy(y_test)

#create dataset and use data loader
testing_samples = utils_data.TensorDataset(x_test, y_test)
data_loader_test = utils_data.DataLoader(testing_samples, batch_size=64,drop_last=False)

#simple model
class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()

        #all the layers
        self.fc1   = nn.Linear(x.shape[1], 20)
        xavier_uniform_(self.fc1.weight.data) #this is how you can change the weight init
        self.drop = nn.Dropout(p=0.5)
        self.fc2   = nn.Linear(20, 1)


    def forward(self, x):
        x = F.relu(self.fc1(x))
        x=  self.drop(x)
        x = self.fc2(x)
        return x

区块 #2

net=Net()

if cuda:
    net.cuda()

# create a stochastic gradient descent optimizer
optimizer = optim.Adam(net.parameters())
# create a loss function (mse)
loss = nn.MSELoss(size_average=False)

# run the main training loop
epochs =20
hold_loss=[]

for epoch in range(epochs):
    cum_loss=0.
    cum_records_epoch =0

    for batch_idx, (data, target) in enumerate(data_loader_trn):
        tr_x, tr_y = data.float(), target.float()
        if cuda:
            tr_x, tr_y = tr_x.cuda(), tr_y.cuda() 

        # Reset gradient
        optimizer.zero_grad()

        # Forward pass
        fx = net(tr_x)
        output = loss(fx, tr_y) #loss for this batch

        cum_loss += output.item() #accumulate the loss

        # Backward 
        output.backward()

        # Update parameters based on backprop
        optimizer.step()

        cum_records_epoch +=len(tr_x)
        if batch_idx % 1 == 0:
            print('Train Epoch: {} [{}/{} ({:.0f}%)]\tLoss: {:.6f}'.format(
            epoch, cum_records_epoch, len(data_loader_trn.dataset),
            100. * (batch_idx+1) / len(data_loader_trn), output.item()))
    print('Epoch average loss: {:.6f}'.format(cum_loss/cum_records_epoch))

    hold_loss.append(cum_loss/cum_records_epoch)  

#training loss
plt.plot(np.array(hold_loss))
plt.show()

Answer 1

可能的原因

不知道"sometimes dramatic differences"是什么，不好回答；但当运行ning [block_1 x1; block_2 x1] xN 时有不同的结果（读“运行ning block_1 然后 block_2 一次；并重复这两个操作 N 次）和[block_1 x1; block_2 xN] x1 是有道理的，因为伪随机数生成器 (PRNG) 和种子是如何工作的。

在第一种情况下，您在每个 block_2 之后重新初始化 block_1 中的 PRNG，因此 block_2 的每个 N 实例都将访问相同的伪随机数序列，由每个 block_1 之前播种。

在第二种情况下，PRNG 仅由单个 block_1 运行初始化一次。所以 block_2 的每个实例都会有不同的随机值。

（有关 PRNG 和种子的更多信息，您可以查看：random.seed(): What does it do?）

简化示例

让我们假设 numpy/CUDA/pytorch 实际上使用了一个非常差的 PRNG，它只有 returns 增量值（即 PRNG(x_n) = PRNG(x_(n-1)) + 1，x_0 = seed）。如果你用 0 作为这个生成器的种子，它将因此 return 1 第一个 random() 调用，2 第二个调用，等等

为了示例，现在让我们也简化您的块：

def block_1():
    seed = 0
    print("seed: {}".format(seed))
    prng.seed(seed)

--

def block_2():
    res = "random results:"
    for i in range(4):
         res  += " {}".format(prng.random())
    print(res)

让我们比较一下 [block_1 x1; block_2 x1] xN 和 [block_1 x1; block_2 xN] x1 与 N=3:

for i in range(3):
    block_1()
    block_2()
# > seed: 0
# > random results: 1 2 3 4
# > seed: 0
# > random results: 1 2 3 4
# > seed: 0
# > random results: 1 2 3 4


block_1()
for i in range(3):
    block_2()
# > seed: 0
# > random results: 1 2 3 4
# > random results: 4 5 6 7
# > random results: 8 9 10 11

可重现的 Pytorch 结果和随机种子

Reproducable Pytorch Results & Random Seeds

python

random

pytorch

random-seed

区块 #1

区块 #2

可能的原因

简化示例