【深度学习】2023李宏毅homework1作业一代码详解
研一刚入门深度学习的小白一枚，想记录自己学习代码的经过，理解每行代码的意思，这样整理方便日后复习也方便理清自己的思路。感觉每天时间都不够用了！！加油啦。
第一部分：导入模块
导入各个模块，代码如下：
# Numerical Operations
import math
import numpy as np
# Reading/Writing Data
import pandas as pd
import os
import csv
# For Progress Bar
from tqdm import tqdm
# Pytorch
import torch
import torch.nn as nn
from torch.utils.data import Dataset, DataLoader, random_split
# For plotting learning curve
from torch.utils.tensorboard import SummaryWriter
在上面程序中，依次导入了
第二部分：切分数据集及预测
随机数，作用是切分训练集和验证集，代码如下：
def same_seed(seed):
'''Fixes random number generator seeds for reproducibility.'''
torch.backends.cudnn.deterministic = True
torch.backends.cudnn.benchmark = False
np.random.seed(seed)
torch.manual_seed(seed)
if torch.cuda.is_available():
torch.cuda.manual_seed_all(seed)
在上面程序中，先调用xxx函数，
接着根据随机数拆分数据集，代码如下：
ef train_valid_split(data_set, valid_ratio, seed):
'''Split provided training data into training set and validation set'''
valid_set_size = int(valid_ratio * len(data_set))
train_set_size = len(data_set) - valid_set_size
train_set, valid_set = random_split(data_set, [train_set_size, valid_set_size], generator=torch.Generator().manual_seed(seed))
return np.array(train_set), np.array(valid_set)
在上面程序中，先调用xxx
接着做预测，下面这段预测程序也作为工具函数，
def predict(test_loader, model, device):
model.eval() # Set your model to evaluation mode.
preds = []
for x in tqdm(test_loader):
x = x.to(device)
with torch.no_grad():
pred = model(x)
preds.append(pred.detach().cpu())
preds = torch.cat(preds, dim=0).numpy()
return preds
在上面程序中，先将模型调成evaluation模式，再设定一个预测结果preds列表，将x
第三部分：数据集
这一部分是数据集，代码如下：
class COVID19Dataset(Dataset):
'''
x: Features.
y: Targets, if none, do prediction.
'''
def __init__(self, x, y=None):
if y is None:
self.y = y
else:
self.y = torch.FloatTensor(y)
self.x = torch.FloatTensor(x)
def __getitem__(self, idx):
if self.y is None:
return self.x[idx]
else:
return self.x[idx], self.y[idx]
def __len__(self):
return len(self.x)
上面这段代码，
第四部分：模型
定义自己的模型，代码如下：
class My_Model(nn.Module):
def __init__(self, input_dim):
super(My_Model, self).__init__()
# TODO: modify model's structure, be aware of dimensions.
self.layers = nn.Sequential(
nn.Linear(input_dim, 16),
nn.ReLU(),
nn.Linear(16, 8),
nn.ReLU(),
nn.Linear(8, 1)
)
def forward(self, x):
x = self.layers(x)
x = x.squeeze(1) # (B, 1) -> (B)
return x
上面这段代码定义了一个继承自nn.Module模块的My_Model类，先在__init__方法中定义层数layers属性，调用nn.Sequential方法列出了5个层，分别是线性层和ReLU层，注意维度分别是input_dim和16，16和8，8和1。接着在forward方法中 得到定义的模型x， 外界可以调用 。
第五部分：特征选择
def select_feat(train_data, valid_data, test_data, select_all=True):
'''Selects useful features to perform regression'''
y_train, y_valid = train_data[:,-1], valid_data[:,-1] # 只需要参考并预测最后一列即可
raw_x_train, raw_x_valid, raw_x_test = train_data[:,:-1], valid_data[:,:-1], test_data # update 1: 去掉第一列 update 2:在特征选择去掉第一列
if select_all:
feat_idx = list(range(raw_x_train.shape[1]))
else:
# update 1: 去掉belief和mental
"""
#feat_idx = [i for i in raw_x_train.shape[1] if i not in ["wbelief_masking_effective", "wbelief_distancing_effective", "wbelief_masking_effective", "worried_finances"]] # update: 不能读取列名，否则array维度不匹配
#feat_idx = [i for i in raw_x_train.shape[1] if i not in [0, 39, 40, 47, 52, 57, 58, 65, 70, 75, 76, 83, 88]] # update: 遍历所有列名，排除不需要的
#feat_idx = [i for i in raw_x_train.shape[1] if i != 0 | i != 39 | i != 40 | i != 47 | i != 52 | i != 57 | i != 58 | i != 65 | i != 70 | i != 75 | i != 76 | i != 83 | i != 88] #update: 整数不可迭代
del_col = [0, 38, 39, 46, 51, 56, 57, 64, 69, 74, 75, 82, 87]
raw_x_train = np.delete(raw_x_train, del_col, axis=1) # update: numpy数组增删查改方法
raw_x_valid = np.delete(raw_x_valid, del_col, axis=1)
raw_x_test = np.delete(raw_x_test, del_col, axis=1)
"""
#update 2:使用前三天的covid like illness和前二天的tested positive cases
get_col = [35, 36, 37, 47, 48, 35+18, 36+18, 37+18, 47+18, 48+18, 35+18*2, 36+18*2, 37+18*2, 47+18*2, 48+18*2, 52, 52+18]
raw_x_train = raw_x_train[:, get_col] # update: numpy数组取某几行某几列
raw_x_valid = raw_x_valid[:, get_col]
raw_x_test = raw_x_test[:, get_col]
return raw_x_train, raw_x_valid, raw_x_test, y_train, y_valid
#feat_idx = [1,1,2,3,4] # TODO: Select suitable feature columns.
return raw_x_train[:,feat_idx], raw_x_valid[:,feat_idx], raw_x_test[:,feat_idx], y_train, y_valid
上面这段代码包含我自己修改的部分，跟着其他大佬的调参步骤更改，加了适当的注释，写在update后面。由列选择得到相应的列…
第六部分：训练
代码如下：
def trainer(train_loader, valid_loader, model, config, device):
criterion = nn.MSELoss(reduction='mean') # Define your loss function, do not modify this.
# Define your optimization algorithm.
# TODO: Please check https://pytorch.org/docs/stable/optim.html to get more available algorithms.
# TODO: L2 regularization (optimizer(weight decay...) or implement by your self).
optimizer = torch.optim.SGD(model.parameters(), lr=config['learning_rate'], momentum=0.9) # update: momentum调整为0.9;
#optimizer = torch.optim.Adam(model.parameters(), lr=config['learning_rate']) # update: 用Adam优化器;
writer = SummaryWriter() # Writer of tensoboard.
if not os.path.isdir('./models'):
os.mkdir('./models') # Create directory of saving models.
n_epochs, best_loss, step, early_stop_count = config['n_epochs'], math.inf, 0, 0
for epoch in range(n_epochs):
model.train() # Set your model to train mode.
loss_record = []
# tqdm is a package to visualize your training progress.
train_pbar = tqdm(train_loader, position=0, leave=True)
for x, y in train_pbar:
optimizer.zero_grad() # Set gradient to zero.
x, y = x.to(device), y.to(device) # Move your data to device.
pred = model(x)
loss = criterion(pred, y)
loss.backward() # Compute gradient(backpropagation).
optimizer.step() # Update parameters.
step += 1
loss_record.append(loss.detach().item())
# Display current epoch number and loss on tqdm progress bar.
train_pbar.set_description(f'Epoch [{epoch+1}/{n_epochs}]')
train_pbar.set_postfix({'loss': loss.detach().item()})
mean_train_loss = sum(loss_record)/len(loss_record)
writer.add_scalar('Loss/train', mean_train_loss, step)
model.eval() # Set your model to evaluation mode.
loss_record = []
for x, y in valid_loader:
x, y = x.to(device), y.to(device)
with torch.no_grad():
pred = model(x)
loss = criterion(pred, y)
loss_record.append(loss.item())
mean_valid_loss = sum(loss_record)/len(loss_record)
print(f'Epoch [{epoch+1}/{n_epochs}]: Train loss: {mean_train_loss:.4f}, Valid loss: {mean_valid_loss:.4f}')
# writer.add_scalar('Loss/valid', mean_valid_loss, step)
if mean_valid_loss < best_loss:
best_loss = mean_valid_loss
torch.save(model.state_dict(), config['save_path']) # Save your best model
print('Saving model with loss {:.3f}...'.format(best_loss))
early_stop_count = 0
else:
early_stop_count += 1
if early_stop_count >= config['early_stop']:
print('\nModel is not improving, so we halt the training session.')
return
上面这段代码…
第七部分：参数
代码如下：
device = 'cuda' if torch.cuda.is_available() else 'cpu'
config = {
'seed': 5201314, # Your seed number, you can pick your lucky number. :)
'select_all': False, # Whether to use all features. update: select_all为False
'valid_ratio': 0.2, # validation_size = train_size * valid_ratio
'n_epochs': 5000, # Number of epochs.
'batch_size': 256,
'learning_rate': 1e-4, # update: 学习率加大为1e-4
'early_stop': 600, # If model has not improved for this many consecutive epochs, stop training.
'save_path': './models/model.ckpt' # Your model will be saved here.
}
上面这部分代码定义了1个设备和8个参数，device是用if-else定义的bool值变量，config用字典表示。
第八部分：开始调用以上定义的方法、对象和参数
数据集处理，代码如下：
same_seed(config['seed'])
train_data, test_data = pd.read_csv('./covid_train.csv').values, pd.read_csv('./covid_test.csv').values # update: .values选中除第一行列名下面的所有行; .values输出的shape一样 (?)
train_data, valid_data = train_valid_split(train_data, config['valid_ratio'], config['seed'])
# Print out the data size.
print(f"""train_data size: {train_data.shape}
valid_data size: {valid_data.shape}
test_data size: {test_data.shape}""")
上面这段代码中，前三行是读入训练和测试的两个.csv文件，得到总的训练集train_data和测试集test_data；再接着对训练集train_data进行切分，得到切分后的训练集train_data和验证集valid_data。
接着进行特征选择，代码如下：
# Select features
x_train, x_valid, x_test, y_train, y_valid = select_feat(train_data, valid_data, test_data, config['select_all'])
# Print out the number of features.
print(f'number of features: {x_train.shape[1]}')
上面这段代码，
接着加载数据，代码如下：
train_dataset, valid_dataset, test_dataset = COVID19Dataset(x_train, y_train), \
COVID19Dataset(x_valid, y_valid), \
COVID19Dataset(x_test)
# Pytorch data loader loads pytorch dataset into batches.
train_loader = DataLoader(train_dataset, batch_size=config['batch_size'], shuffle=True, pin_memory=True)
valid_loader = DataLoader(valid_dataset, batch_size=config['batch_size'], shuffle=True, pin_memory=True)
test_loader = DataLoader(test_dataset, batch_size=config['batch_size'], shuffle=False, pin_memory=True)
上面这段代码，train和valid的dataset进行了shuffle，而test的dataset不需要shuffle。
接着进行训练，代码如下：
model = My_Model(input_dim=x_train.shape[1]).to(device) # put your model and data on the same computation device.
trainer(train_loader, valid_loader, model, config, device)
上面这段代码，
接着进行预测，并保存预测结果，代码如下：
def save_pred(preds, file):
''' Save predictions to specified file '''
with open(file, 'w') as fp:
writer = csv.writer(fp)
writer.writerow(['id', 'tested_positive'])
for i, p in enumerate(preds):
writer.writerow([i, p])
model = My_Model(input_dim=x_train.shape[1]).to(device)
model.load_state_dict(torch.load(config['save_path'])) # update: tensor size mismatch，所以暂时先注释掉
preds = predict(test_loader, model, device)
save_pred(preds, 'pred.csv')
上面这段代码先定义了一个save_pred方法，调用open创建一个.csv文件…