It's used often in this project, even in the loss. But I cannot find any documentation about this value. https://github.com/Microsoft/DualLearning/blob/1068687f839bb05b01c69833b2852cdbf488a2c2/DSL_ImgProcess/data2/cifar10_data.py#L80-L85 https://github.com/Microsoft/DualLearning/blob/6c7873747400359bd0f0aa69dabe23b814ea2f0f/DSL_ImgProcess/monitor.py#L399-L400 https://github.com/Microsoft/DualLearning/blob/6c7873747400359bd0f0aa69dabe23b814ea2f0f/DSL_ImgProcess/monitor.py#L152-L157 Does it mean the probability of the image?
It's used often in this project, even in the loss. But I cannot find any documentation about this value.
https://github.com/Microsoft/DualLearning/blob/1068687f839bb05b01c69833b2852cdbf488a2c2/DSL_ImgProcess/data2/cifar10_data.py#L80-L85
https://github.com/Microsoft/DualLearning/blob/6c7873747400359bd0f0aa69dabe23b814ea2f0f/DSL_ImgProcess/monitor.py#L399-L400
https://github.com/Microsoft/DualLearning/blob/6c7873747400359bd0f0aa69dabe23b814ea2f0f/DSL_ImgProcess/monitor.py#L152-L157
Does it mean the probability of the image?