pytorch Dataset 데이터 세트와 Dataloader 교체 데이터 세트

1123 단어
import torch
from torch.utils.data import Dataset,DataLoader

class SmsDataset(Dataset):

    def __init__(self):
        self.file_path = "./SMSSpamCollection"
        self.lines = open(self.file_path,encoding="utf-8").readlines()

    def __getitem__(self, index):
        line = self.lines[index].strip()
        label = line.split("\t")[0]
        sent = line.split("\t")[1]
        return label,sent

    def __len__(self):
        return len(self.lines)

sms_dataset = SmsDataset()
dataloader = DataLoader(sms_dataset,batch_size=2,shuffle=True)


if __name__ == '__main__':
    for idx,(label,sent) in enumerate(dataloader):
        print(idx)
        print(label)
        print(sent)
        break
    print(len(sms_dataset))
    print(len(dataloader))

  
0
('ham', 'spam')
('And popping  ibuprofens was no help.', 'This is the 2nd time we have tried 2 contact u. U have won the 750 Pound prize. 2 claim is easy, call 08712101358 NOW! Only 10p per min. BT-national-rate')
5574
2787

좋은 웹페이지 즐겨찾기