lzAllen的博客

初次跑通MMDetection的MaskRCNN Demo 惯例先跑一下demo，新建一个HellowMMDection.py，将这个脚本放在./demo中去: # Check Pytorch installation import torch, torchvision print(torch.__version__, torch.cuda.is_available()) # Check MMDetection installation import mmdet print(mmdet.__ver

标签平滑的pytorch的简单实现 import torch import torch.nn as nn def CrossEntropyLoss_label_smooth(outputs, targets, num_classes=10, epsilon=0.1): N = targets.size(0) # torch.Size([8, 10]) # 初始化一个矩阵, 里面的值都是epsilon / (num_classes - 1) smoothed_labels