Open ggghamd opened 2 years ago
The main.py is not a complete example code, it's some code snippets. You shoud implement your own "custom data loader", refer to torchvision.datasets.CIFAR10. Your error is caused by null implementation of CustomDataset.
nnt@mycomp:~/pytorch-tutorial/tutorials/01-basics/pytorch_basics$ python3 main.py
tensor(2.)
tensor(1.)
tensor(1.)
w: Parameter containing:
tensor([[-0.0246, 0.4567, -0.4118],
[ 0.1679, -0.5086, -0.3876]], requires_grad=True)
b: Parameter containing:
tensor([0.0529, 0.2922], requires_grad=True)
loss: 1.949414610862732
dL/dw: tensor([[-0.7870, 1.5242, -0.0795],
[ 0.8206, -0.9073, -0.7711]])
dL/db: tensor([-0.5564, 0.0907])
loss after 1 step optimization: 1.896445631980896
Files already downloaded and verified
torch.Size([3, 32, 32])
6
Traceback (most recent call last):
File "/home/nnt/pytorch-tutorial/tutorials/01-basics/pytorch_basics/main.py", line 124, in <module>
images, labels = data_iter.next()
AttributeError: '_SingleProcessDataLoaderIter' object has no attribute 'next'
Also this
https://github.com/yunjey/pytorch-tutorial/blob/master/tutorials/01-basics/pytorch_basics/main.py: