[人工智能]文本分类半监督学习问题（十三）

2021SC@SDUSC

最后对train.py中的main函数进行分析：

首先是读取数据并且建立数据加载器

def main():
    global best_acc
    train_labeled_set, train_unlabeled_set, val_set, test_set, n_labels = get_data(
        args.data_path, args.n_labeled, args.un_labeled, model=args.model, train_aug=args.train_aug)
    labeled_trainloader = Data.DataLoader(
        dataset=train_labeled_set, batch_size=args.batch_size, shuffle=True)
    unlabeled_trainloader = Data.DataLoader(
        dataset=train_unlabeled_set, batch_size=args.batch_size_u, shuffle=True)
    val_loader = Data.DataLoader(
        dataset=val_set, batch_size=512, shuffle=False)
    test_loader = Data.DataLoader(
        dataset=test_set, batch_size=512, shuffle=False)

定义模型并且定义优化器

    model = MixText(n_labels, args.mix_option).cuda()
    model = nn.DataParallel(model)
    optimizer = AdamW(
        [
            {"params": model.module.bert.parameters(), "lr": args.lrmain},
            {"params": model.module.linear.parameters(), "lr": args.lrlast},
        ])

开始训练：

 for epoch in range(args.epochs):

        train(labeled_trainloader, unlabeled_trainloader, model, optimizer,
              scheduler, train_criterion, epoch, n_labels, args.train_aug)
 val_loss, val_acc = validate(
            val_loader, model, criterion, epoch, mode='Valid Stats')

        print("epoch {}, val acc {}, val_loss {}".format(
            epoch, val_acc, val_loss))

        if val_acc >= best_acc:
            best_acc = val_acc
            test_loss, test_acc = validate(
                test_loader, model, criterion, epoch, mode='Test Stats ')
            test_accs.append(test_acc)
            print("epoch {}, test acc {},test loss {}".format(
                epoch, test_acc, test_loss))

        print('Epoch: ', epoch)

        print('Best acc:')
        print(best_acc)

        print('Test acc:')
        print(test_accs)

    print("Finished training!")
    print('Best acc:')
    print(best_acc)

    print('Test acc:')
    print(test_accs)