Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

use different encoder #61

Open
thesuncool opened this issue Jan 13, 2023 · 3 comments
Open

use different encoder #61

thesuncool opened this issue Jan 13, 2023 · 3 comments

Comments

@thesuncool
Copy link

hello,i use blip model replace clip model to extract image feauture, but after evaluation,the following problems arise, the resulting sentences are all the same sentence.
image

@rongtongxueya
Copy link

I've had the same problem. It's the same thing over and over again. Have you found the reason for this problem

@thesuncool
Copy link
Author

thesuncool commented Jul 11, 2023 via email

@rongtongxueya
Copy link

没有,而且我发现复现这篇代码,得出的评分很低,达不到文章中的效果,但是具体差多少我忘了,你跑出的结果可以给我看一下吗

------------------ 原始邮件 ------------------ 发件人: @.>; 发送时间: 2023年7月11日(星期二) 晚上8:51 收件人: @.>; 抄送: @.>; @.>; 主题: Re: [rmokady/CLIP_prefix_caption] use different encoder (Issue #61) 我也遇到过同样的问题。一遍又一遍的重复着同样的事情。你找到这个问题的原因了吗 — 直接回复此邮件,在GitHub上查看,或者取消订阅. @.***与>.
我倒是训练了,但是我刚开始肝这个,还没弄明白怎么输出这个BLEU的这些分数呢,你能不能给我点建议教教我啊

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

2 participants