-
Notifications
You must be signed in to change notification settings - Fork 0
/
text_classification.py
64 lines (58 loc) · 2.33 KB
/
text_classification.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
import torch
from transformers import BertTokenizerFast
from utils import SpamText
from networks import ManyToOne
from torch.optim import AdamW
from torch.utils.data import DataLoader
import torch.nn as nn
from torch.optim.lr_scheduler import MultiStepLR
def run():
tokenizer = BertTokenizerFast.from_pretrained("bert-base-uncased")
model= ManyToOne(vocab_size=tokenizer.vocab_size,\
input_size=1000,\
hidden_size=128,\
output_size=2)
#train
device = torch.device("cuda:0" if torch.cuda.is_available() else "cpu")
model.to(device)
#hyp
epoches = 30
lr=0.01
batch_size=128
train_dataloader=DataLoader(SpamText("datasets/text_classification/SPAM-text-message_20170820_Data_train.csv",tokenizer),\
batch_size,
shuffle=True)
optimizer=AdamW(model.parameters(),lr=lr,weight_decay=1e-5)
criteria=nn.CrossEntropyLoss()
schedule=MultiStepLR(optimizer,milestones=[int(epoches*0.6),int(epoches*0.8)], gamma=0.1)
for e in range(epoches):
running_loss = 0.0
for i,b in enumerate(train_dataloader):
optimizer.zero_grad()
logits=model(b[0][0].to(device))
loss = criteria(logits,b[1].to(device))
loss.backward()
optimizer.step()
running_loss += loss.item()
if i % 10 == 9: # print average loss every 10 mini batches
print('[Epoch:%d, %5d / %5d] loss: %.6f lr: %.6f' %
(e + 1, i + 1, len(train_dataloader),running_loss / 10,schedule.get_last_lr()[0]))
running_loss = 0.0
schedule.step()
#test
test_dataloader=DataLoader(SpamText("datasets/text_classification/SPAM-text-message_20170820_Data_test.csv",tokenizer),\
batch_size)
with torch.no_grad():
model.eval()
total=0
correct=0
for i,b in enumerate(test_dataloader):
logits=model(b[0][0].to(device))
pred=logits.softmax(dim=1).argmax(dim=1).data
target=b[1].to(device).argmax(dim=1).data
_=(pred==target)
correct+=_.sum().item()
total+=len(_)
print("test acc:",correct/total)
if __name__ == "__main__":
run()