CNN半监督学习中的问题

Question

我进行了半监督学习，在数据集中标注未标注的图像。CNN模型以无标号图像为输入，经过softmax计算，生成一个概率指数。如果值超过某个数字(例如0.65)，我将标记图像并将其添加到火车组中。获取说服力数据集的代码：

def get_pseudo_labels(trainset, dataset, model, threshold=0.65):
# This functions generates pseudo-labels of a dataset using given model.
# It returns an instance of DatasetFolder containing images whose prediction confidences exceed a given threshold.
device = "cuda" if torch.cuda.is_available() else "cpu"

# Construct a data loader.
data_loader = DataLoader(dataset, batch_size=batch_size, shuffle=False)
# The dataset is unlabelled image

# Make sure the model is in eval mode.
model.eval()
# Define softmax function.
softmax = nn.Softmax(dim=-1)

# Iterate over the dataset by batches.
for batch in tqdm(data_loader):

    img, labels = batch
    # Forward the data
    # Using torch.no_grad() accelerates the forward process.
    with torch.no_grad():
        logits = model(img.to(device))

    # Obtain the probability distributions by applying softmax on logits.
    probs = softmax(logits)
    # calculate probs

    for j in range(0, batch_size):
        for i in range(0, 11):
            if probs[j][i].item() > threshold:
                batch[1][j] = torch.Tensor([i]) # Label the imgae
                temp = batch[0][j] + batch[1][j] # contact two tensor
                trainset = ConcatDataset([trainset, temp]) # add this labelled image into trainset

model.train()
return trainset

编者提醒我：

if probsj.item() >阈值: IndexError:索引2对于尺寸为2的维度0是超出界限的

不过，我可以正常打印问题。

        for j in range(0, batch_size):
        for i in range(0, 11):
            print('batch:', j)
            print('The value of label', i)
            print(probs[j][i])
            if probs[j][i].item() > threshold:
                batch[1][j] = torch.Tensor([i])
                temp = batch[0][j] + batch[1][j]
                trainset = ConcatDataset([trainset, temp])

输出：

...
batch: 63
The value of label 9
tensor(0.0859, device='cuda:0')
batch: 63
The value of label 10
tensor(0.0977, device='cuda:0')

我不知道IndexError是什么意思..。

国际管理小组的主要工作是：

tensor([...(img)],[...(label)])

Answer 1

确保dataset % batch_size = 0。

对于batch_size (例如4)，您应该在数据集中有很多示例(8或12或16等等)。

这里16 %4=0