pytorch 用sampler划分数据集

train_dataloader = DataLoader(dataset=dataset,batch_size=4,sampler=sampler.SequentialSampler(range(24)))
print(len(train_dataloader))
val_dataloader = DataLoader(dataset=dataset,batch_size=4,sampler=sampler.SequentialSampler(range(24,28)))
print(len(val_dataloader))
输出
6
1

文件中有28个数据 然后

sampler=sampler.SequentialSampler(range(24))

采样前24个 batchsize是4 所以 dataloader长度为6