Pytorch 1.12
构建embedding 层：
import torch,torchtext,torchdata
torch.nn.Embedding(num_embeddings = 1000,embedding_dim =2)
Embedding(1000, 2)
train_iter, test_iter = torchtext.datasets.IMDB()
print(next(iter(train_iter)))
('neg', 'I rented I AM CURIOUS-YELLOW from my video store because of all the controversy that surrounded it when it was first released in 1967. I also heard that at first it was seized by U.S. customs if it ever tried to enter this country, therefore being a fan of films considered "controversial" I really had to see this for myself.<br /><br />The plot is centered around a young Swedish drama student named Lena who wants to learn everything she can about life. In particular she wants to focus her attentions to making some sort of documentary on what the average Swede thought about certain political issues such as the Vietnam War and race issues in the United States. In between asking politicians and ordinary denizens of Stockholm about their opinions on politics, she has sex with her drama teacher, classmates, and married men.<br /><br />What kills me about I AM CURIOUS-YELLOW is that 40 years ago, this was considered pornographic. Really, the sex and nudity scenes are few and far between, even then it\'s not shot like some cheaply made porno. While my countrymen mind find it shocking, in reality sex and nudity are a major staple in Swedish cinema. Even Ingmar Bergman, arguably their answer to good old boy John Ford, had sex scenes in his films.<br /><br />I do commend the filmmakers for the fact that any sex shown in the film is shown for artistic purposes rather than just to shock people and make money to be shown in pornographic theaters in America. I AM CURIOUS-YELLOW is a good film for anyone wanting to study the meat and potatoes (no pun intended) of Swedish cinema. But really, this film doesn\'t have much of a plot.')
unique_labels=set([label for (label, text) in train_iter])
print(unique_labels)
num_class = len(unique_labels)
{'neg', 'pos'}
from torchtext.data.utils import get_tokenizer
from torchtext.vocab import build_vocab_from_iterator
tokenizer = get_tokenizer('basic_english')
print(tokenizer('this is a book about PyTorch.'))
['this', 'is', 'a', 'book', 'about', 'pytorch', '.']
def yield_tokens(data):
for _,text in data:
yield tokenizer(text)
vocab = build_vocab_from_iterator(
yield_tokens(train_iter),
specials=["<pad>","<unk>"],
min_freq=5
)
print(len(vocab))
30123
vocab.set_default_index(vocab["<unk>"])
print(vocab(["this","is","a","book","about","pytorch","/"]))
[14, 10, 6, 276, 50, 1, 2192]
text_pipeline = lambda x: vocab(tokenizer(x))
label_pipeline = lambda x: int(x=="pos")
text_pipeline('This is a book about PyTorch.')
print(label_pipeline('pos'))
from torchtext.data.functional import to_map_style_dataset
train_dataset = to_map_style_dataset(train_iter)
test_dataset = to_map_style_dataset(test_iter)
from torch.utils.data import DataLoader
device = torch.device('cuda' if torch.cuda.is_available() else 'cpu')
# 批次处理
def collate_batch(batch):
label_list, text_list, offsets = [],[],[0]
for (_label, _text) in batch:
label_list.append(label_pipeline(_label))
process_text = torch.tensor(text_pipeline(_text),dtype=torch.int64)
text_list.append(process_text)
offsets.append(process_text.size(0))
label_list = torch.tensor(label_list)
text_list = torch.cat(text_list)
offsets = torch.tensor(offsets[:-1]).cumsum(dim=0)
return label_list.to(device), text_list.to(device), offsets.to(device)
BATCHSIZE=64
train_dataloader = DataLoader(train_dataset, batch_size=BATCHSIZE,shuffle=True, collate_fn=collate_batch)
test_dataloader = DataLoader(test_dataset, batch_size=BATCHSIZE, collate_fn=collate_batch)
label_list, text_list, offsets = next(iter(train_dataloader))
len(label_list), len(text_list), len(offsets)
import torch.nn as nn
vocab_size = len(vocab) # 获取词表大小
embedding_dim = 100 # 定义词嵌入向量大小
class TextClassificationModel(nn.Module):
def __init__(self, vocab_size, embed_dim, num_class) -> None:
super().__init__()
self.embedding = nn.EmbeddingBag(vocab_size, embed_dim,sparse=True)
self.fc = nn.Linear(embed_dim, num_class)
# 尼玛对上了，在这里呢，初始化权重
self.init_weights()
def init_weights(self):
initrange = 0.5
self.embedding.weight.data.uniform_(-initrange,initrange)
self.fc.weight.data.uniform_(-initrange, initrange)
self.fc.bias.data.zero_()
def forward(self, text, offsets):
embedded = self.embedding(text, offsets)
return self.fc(embedded)
vocab_size = len(vocab)
model=TextClassificationModel(vocab_size, embedding_dim, num_class).to(device)
model
TextClassificationModel(
(embedding): EmbeddingBag(30123, 100, mode=mean)
(fc): Linear(in_features=100, out_features=2, bias=True)
)
loss_fn = nn.CrossEntropyLoss() # 分类问题的损失函数
from torch.optim import lr_scheduler # 用于对学习速率做衰减
optimizer = torch.optim.SGD(model.parameters(), lr=0.1)
# 定义学习速率衰减策略
exp_lr_scheduler = lr_scheduler.StepLR(optimizer,step_size=20,gamma=0.1)
def train(dataloader):
total_acc,total_count,total_loss = 0,0,0
model.train() # 这是啥意思，训练的时候开启BN和Dropout，
for label,text, offsets in dataloader:
# @todo为什么返回的是三项
predicted_label=model(text,offsets) #模型调用时要输入offsets
loss = loss_fn(predicted_label, label)
# 反向传播
optimizer.zero_grad()
loss.backward()
optimizer.step()
with torch.no_grad():
total_acc += (predicted_label.argmax(1)==label).sum().item()
total_count += label.size(0)
total_loss += loss.item() * label.size(0)
return total_loss/total_count, total_acc/total_count
def test(dataloader):
model.eval()
total_acc, total_count, total_loss = 0,0,0
with torch.no_grad():
for idx,(label,text,offsets) in enumerate(dataloader):
predicted_label = model(text, offsets)
loss = loss_fn(predicted_label,label)
total_acc += (predicted_label.argmax(1)==label).sum().item()
total_count += label.size(0)
total_loss += loss.item() * label.size(0)
return total_loss/total_count, total_acc/total_count
def fit(epochs, train_dl, test_dl):
train_loss = []
train_acc = []
test_loss = []
test_acc = []
for epoch in range(epochs):
epoch_loss, epoch_acc = train(train_dl)
epoch_test_loss, epoch_test_acc = test(test_dl)
train_loss.append(epoch_loss)
train_acc.append(epoch_acc)
test_loss.append(epoch_test_loss)
test_acc.append(epoch_test_acc)
exp_lr_scheduler.step()
template =("epoch:{:2d}, train_loss:{:.5f},train_acc:{:.1f}%, test_loss:{:.5f}, test_acc:{:.1f}%")
print(template.format(epoch,epoch_loss,epoch_acc*100, epoch_test_loss,epoch_test_acc*100))
print("Done!")
return train_loss, test_loss, train_acc, test_acc
epochs = 20
fit(epochs, train_dataloader, test_dataloader)
epoch: 0, train_loss:0.69063,train_acc:53.0%, test_loss:0.68775, test_acc:54.7%
Done!
epoch: 1, train_loss:0.68182,train_acc:58.4%, test_loss:0.68003, test_acc:59.2%
Done!
epoch: 2, train_loss:0.67389,train_acc:61.6%, test_loss:0.67334, test_acc:60.2%
Done!
epoch: 3, train_loss:0.66643,train_acc:63.2%, test_loss:0.66645, test_acc:62.6%
Done!
epoch: 4, train_loss:0.65892,train_acc:64.4%, test_loss:0.65915, test_acc:63.5%
Done!
epoch: 5, train_loss:0.65146,train_acc:65.2%, test_loss:0.65190, test_acc:65.0%
Done!
epoch: 6, train_loss:0.64365,train_acc:66.1%, test_loss:0.64442, test_acc:65.7%
Done!
epoch: 7, train_loss:0.63593,train_acc:66.8%, test_loss:0.63676, test_acc:66.6%
Done!
epoch: 8, train_loss:0.62797,train_acc:67.7%, test_loss:0.62896, test_acc:67.3%
Done!
epoch: 9, train_loss:0.61967,train_acc:68.7%, test_loss:0.62102, test_acc:68.3%
Done!
epoch:10, train_loss:0.61128,train_acc:69.3%, test_loss:0.61303, test_acc:69.1%
Done!
epoch:11, train_loss:0.60308,train_acc:70.2%, test_loss:0.60502, test_acc:69.7%
Done!
epoch:12, train_loss:0.59478,train_acc:70.9%, test_loss:0.59658, test_acc:70.5%
...
epoch:18, train_loss:0.54764,train_acc:74.2%, test_loss:0.55076, test_acc:73.6%
Done!
epoch:19, train_loss:0.54041,train_acc:74.9%, test_loss:0.54327, test_acc:74.3%
Done!
Output is truncated. View as a scrollable element or open in a text editor. Adjust cell output settings...
434
被折叠的 条评论
为什么被折叠?
到【灌水乐园】发言
