pytorch实现吴恩达机器学习课程实验四（反向传播神经网络）

import torch
import numpy as np
# import torch.utils.data as datasets
from torch.utils.data import Dataset
from torch.utils.data import DataLoader
from torch import nn
from scipy.io import loadmat

learning_rate = 0.01
epoch_num = 5

data = loadmat('/content/drive/MyDrive/colab/data/NUMIMG/ex4data1.mat')
print(data.keys())
print(len(data['X']),type(data['X']),data['X'].shape)
print(np.max(data['y']))

class ImagDataset(Dataset):
  def __init__(self):
    path = '/content/drive/MyDrive/colab/data/NUMIMG/ex4data1.mat'
    data = loadmat(path)
    self.data_x = torch.from_numpy(data['X']).float()
    self.data_y = torch.from_numpy(data['y']).long().reshape(-1) - 1
    self.sample_num = len(data['X'])
  def __getitem__(self,index):
    return self.data_x[index], self.data_y[index]
  def __len__(self):
    return self.sample_num

dataset = ImagDataset()
dataloader = DataLoader(dataset=dataset, batch_size=32,shuffle=True)


class NeuralNet(nn.Module):
  def __init__(self):
    super(NeuralNet, self).__init__()
    self.linear1 = nn.Linear(400, 512)
    self.relu = nn.ReLU()
    self.linear2 = nn.Linear(512, 10)
  def forward(self, x):
    out = self.linear1(x)
    out = self.relu(out)
    out = self.linear2(out)
    return out

model = NeuralNet()

loss_fc = nn.CrossEntropyLoss()
optim = torch.optim.Adam(model.parameters(), lr=learning_rate)

total_step = len(dataloader)
print(total_step)
for epoch in range(epoch_num):
  for i, (images, labels) in enumerate(dataloader):
    out = model(images)

    # print(out.shape,labels.shape)
    loss = loss_fc(out,labels)
    
    optim.zero_grad()
    loss.backward()
    optim.step()

    if (i + 1) % 32 == 0:
      print('epoch [{}/{}] step [{}/{}] loss: {:.4f}'.format(epoch + 1, epoch_num, i + 1, total_step, loss.item()))

with torch.no_grad():
  correct = 0
  total = 0
  for images,label in dataloader:
    out = model(images)
    val, pred = torch.max(out, 1)
    total += images.shape[0]
    correct += (pred == label).sum().item()
  print('accuracy on {} test cases is {}%'.format(len(dataset), 100 * correct / len(dataset)))

里面有很多细节，比如在colab使用本地数据，并结合dataset和dataloader使用；dataset的一些细节问题