Fast-AutoAug-Torch
Fast-AutoAug-Torch copied to clipboard
Published
20 hours ago •
zhanghang1989
zhanghang1989
Results
1
Fast-AutoAug-Torch issues
Sort by
recently updated
recently updated
newest added
[第145行](https://github.com/zhanghang1989/Fast-AutoAug-Torch/blob/master/search_policy.py#L145)` transform_val.transforms.insert(0, Augmentation(auto_policy))`中插入的`auto_policy`为二维数组，这就造成在augment.py文件中[第55行](https://github.com/zhanghang1989/Fast-AutoAug-Torch/blob/a991a4ee7f58c556ad00d50ff315992bb6adf535/augment.py#L55) `for name, pr, level in policy`中的`policy`实际为一维数组，导致程序无法继续运行。 改为` transform_val.transforms.insert(0, Augmentation([auto_policy]))`可以解决这个问题。
