About the results of APTOS2019

[kjRainy]

Hello, thank you very much for your research contribution. I trained the model based on the data partition of APTOS2019 and the configuration parameters in aptos.yml that you provided. Why is the result lower than the original text? Are there any other configurations that are not correct? Looking forward to your reply, thank you very much.

Here are my results:
[Test:] Average accuracy: 81.232, F1: 0.653

[xyz1110]

Hello, thank you very much for your research contribution. I trained the model based on the data partition of APTOS2019 and the configuration parameters in aptos.yml that you provided. Why is the result lower than the original text? Are there any other configurations that are not correct? Looking forward to your reply, thank you very much.

Here are my results: [Test:] Average accuracy: 81.232, F1: 0.653
Hi, I've recently been working on the reproduction of this project as well and have some questions for you ，1225622135@qq.com

[zbzzz]

Hello, I noticed that the lambda_ce item in the configuration file is commented out. Should it be uncommented during normal training?

[IG71Y]

Hello, I noticed that the lambda_ce item in the configuration file is commented out. Should it be uncommented during normal training?

Hi, did you uncomment the "#lambda_ce: 0.01"item during normal training? What is the effect?