我们使用PyTorch实现了我们的网络。We adopt ResNet-50 pre-trained from ImageNet as backbone。与transunet一致我们训练我们的模型使用SDG 优化器with基础的学习率为001动量为09weight_decay为00001。我们使用同样的学习率下降策略：翻译成英文

We implemented our network using PyTorch. We used ResNet-50 pre-trained from ImageNet as the backbone. Similar to TransUNet, we trained our model using the SDG optimizer with a base learning rate of 0.01, momentum of 0.9, and weight decay of 0.0001. We used the same learning rate decay strategy.