search

ZhaoZhibin / UDTL

Source codes for the paper "Applications of Unsupervised Deep Transfer Learning to Intelligent Fault Diagnosis: A Survey and Comparative Study" published in TIM
MIT License
400 stars 103 forks source link

保存最佳模型参数代码的或许有用的建议 #19

Open zjq-learn opened 4 months ago

zjq-learn commented 4 months ago

train_utils_combines文件中的torch.save代码应该放到所有epoch结束之后。torch.save前面的if循环条件中or后面的循环条件应该去除。