Qingyuan Zheng
Qingyuan Zheng
We get tiles by kernel score, but we don't seem to get a satisfactory answer. How to solve this. 
我在推理的时候,输出的预测结果shape不对,按理说应该是(64,128,128),结果却是(1,64,128,128),错误的npy大小约为正常预测结果的两倍。这是什么原因? --- 欢迎您反馈PaddleSeg使用问题,辛苦您提供以下信息,方便我们快速定位和解决问题: 1. PaddleSeg版本:(PaddleSeg release/2.3) 2. PaddlePaddle版本:(PaddlePaddle 2.3.0) 3. 操作系统信息:(Linux) 4. Python版本号:(python3.7) 5. CUDA/cuDNN版本:( CUDA10.2) 6. 其他内容: (增加其他与问题相关的内容)
Help
superpatch_network_constrution.py doesn't seem to work, are you sure root_dir is /Graph_test/?