[FIXED] Loss function error in pytorch-How should I use the loss function?

Issue

I wrote this code:

import torch
import torch.nn as nn
import torch.nn.functional as f
import math
import matplotlib.pyplot as plt
import numpy as np
import random

class dnn(nn.Module):
    def __init__(self):
        super(dnn, self).__init__()
        self.l1 = nn.Linear(in_features=1, out_features=3)
        self.l2 = nn.Linear(in_features=3, out_features=3)
        self.l3 = nn.Linear(in_features=3, out_features=1)

    def forward(self, out):
        out = f.leaky_relu(self.l1(out))
        out = f.leaky_relu(self.l2(out))
        out = f.leaky_relu(self.l3(out))
        return out

model = dnn()
criterion = nn.CrossEntropyLoss()
optimizer = torch.optim.Adam(model.parameters(), lr=0.1)

for k in range(100):
    for i in range(30):
        outputs = model(torch.tensor([i]).type(torch.FloatTensor))
        loss = criterion(outputs, torch.tensor([math.cos(i)]).type(torch.FloatTensor))
        *print(loss)*
        loss.backward()
        optimizer.zero_grad()
        optimizer.step()
    print('{}%'.format(k))

x = []
y = []
p = []

for i in range(30):
    x.append(i)
    y.append(math.cos(i))
    p.append(str(model(torch.tensor([i]).type(torch.FloatTensor))).split('[')[1].split(']')[0])

plt.plot(x, p, 'g')
plt.plot(x, y, 'b')

plt.show()

However, the results predicted by this model were very strange.

https://i.stack.imgur.com/bs6Q2.png

So I tried a few more times and kept getting similar results. While I was looking at various variables in the code to solve the problem, I found something strange.

print(loss):

tensor(-0., grad_fn=<DivBackward1>)
tensor(-0., grad_fn=<DivBackward1>)
tensor(-0., grad_fn=<DivBackward1>)
tensor(-0., grad_fn=<DivBackward1>)
tensor(-0., grad_fn=<DivBackward1>)
tensor(-0., grad_fn=<DivBackward1>)
tensor(-0., grad_fn=<DivBackward1>)
tensor(-0., grad_fn=<DivBackward1>)
tensor(-0., grad_fn=<DivBackward1>)
tensor(-0., grad_fn=<DivBackward1>)
tensor(-0., grad_fn=<DivBackward1>)
tensor(-0., grad_fn=<DivBackward1>)
tensor(-0., grad_fn=<DivBackward1>)
tensor(-0., grad_fn=<DivBackward1>)
tensor(-0., grad_fn=<DivBackward1>)
tensor(-0., grad_fn=<DivBackward1>)

The loss function has not changed. How do I solve this?

Solution

you need loss function for task of regression, for example

criterion = nn.L1Loss()

nn.CrossEntropyLoss() expects targets as indexes of class or probabilities of class, you can change your network for binary classification task

import torch
import torch.nn as nn
import torch.nn.functional as f
import math
import matplotlib.pyplot as plt
import numpy as np
import random

class dnn(nn.Module):
    def __init__(self):
        super(dnn, self).__init__()
        self.l1 = nn.Linear(in_features=1, out_features=3)
        self.l2 = nn.Linear(in_features=3, out_features=3)
        self.l3 = nn.Linear(in_features=3, out_features=2)

    def forward(self, out):
        out = f.leaky_relu(self.l1(out))
        out = f.leaky_relu(self.l2(out))
        out = f.leaky_relu(self.l3(out))
        
        return out

model = dnn()
criterion = nn.CrossEntropyLoss()
optimizer = torch.optim.Adam(model.parameters(), lr=0.1)

for k in range(100):
    for i in range(30):
        # print(torch.tensor([i]))
        outputs = model(torch.tensor([i]).type(torch.FloatTensor))
        
        loss = criterion(outputs, torch.tensor([math.cos(i), 0]).type(torch.FloatTensor))
        loss.backward()
        optimizer.zero_grad()
        optimizer.step()
    print('{}%'.format(k))

x = []
y = []
p = []

for i in range(30):
    x.append(i)
    y.append(math.cos(i))
    p.append(str(model(torch.tensor([i]).type(torch.FloatTensor))).split('[')[1].split(']')[0])

plt.plot(x, p, 'g')
plt.plot(x, y, 'b')

plt.show()

Answered By - Anna Andreeva Rogotulka