Pytorch study website:
https://pytorch.org/tutorials/beginner/deep_learning_60min_blitz.html
https://pytorch.org/
https://morvanzhou.github.io/tutorials/machine-learning/torch/
$ start anaconda prompt by adminstrator
$ conda install pytorch torchvision cudatoolkit=10.0 -c pytorch
Windows Know issues:
Q: File "C:\ProgramData\Anaconda3\lib\multiprocessing\reduction.py", line 60, in dump
ForkingPickler(file, protocol).dump(obj)
BrokenPipeError: [Errno 32] Broken pipe
A: Windows OS have problem about multi-thread load data, so you just need to set num_workers=0
For example:
torch.utils.data.DataLoader(trainset, batch_size=4, shuffle=True, num_workers=0)
$ pip3 install torch torchvision
pytorch train
refer: pytorch_learn/train_cnn_cifar10/train.py
pytorch test
refer: pytorch_learn/train_cnn_cifar10/test.py
cpp inference
refer: pytorch_learn/test_cifar10_cpp
detail refer:pytorch_learn/test_cifar10_cpp/readme.md