Open Addlien opened 1 year ago
Hi, not sure why the problem occurs, all my tries are also collapsing to label 3.
你好,我尝试run你的程序,但是在train.py里,我将每一个的预测结果-->> 还是中文简体好用
Hello, I did not find the weight magnet.pt. In addition, please help me how to generate the NPZ file. thank you!
Excuse me, have you solved this problem yet? I encountered the same issue. T T
Sorry, I cannot solve it,and give up for a long time.Good luck for you.
Add+ @.***
------------------ 原始邮件 ------------------ 发件人: "Justin900429/Facial-Graph-Representation-Learning" @.>; 发送时间: 2023年5月12日(星期五) 中午12:05 @.>; @.**@.>; 主题: Re: [Justin900429/Facial-Graph-Representation-Learning] hi, there comes some problems in running your code (Issue #6)
Excuse me, have you solved this problem yet? I encountered the same issue. T T
— Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: @.***>
Try initialising the weights properly. (Sorry for not testing this before)
妳好,我嘗試run妳的程式,但是在train.py 裏,我將每壹個epoch trainlist的預測結果打印出來,發現最後都會變成3:other, 請問這是為什麼呢? 該如何解抉?
Hi, I try to run your project, but when I print train_list in every epoch in
python.py
, only to find that all predicted labels become '3'--other gradually after at most three epoches. Can you tell me why and how to fix it.