changlin31 / DNA

(CVPR 2020) Block-wisely Supervised Neural Architecture Search with Knowledge Distillation
234 stars 35 forks source link

train_pipeline.yaml 默认值 #24

Closed nbl97 closed 3 years ago

nbl97 commented 3 years ago

您好。训练超网时 train_pipeline.yaml 某些参数的值与注释中的default值不一致,实际训练时该用哪个值呢?比如 opt、re_prob、warmup_epochs等。十分感谢

changlin31 commented 3 years ago

使用目前代码里的就好