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.

Sign up for GitHub

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

Jump to bottom

[Question]: 想问下AquilaDecoderLayer里的传参为什么是float的？这样是用的float类型的attention #554

Open
Chenjingliang1 opened this issue Nov 7, 2023 · 2 comments
Open

[Question]: 想问下AquilaDecoderLayer里的传参为什么是float的？这样是用的float类型的attention #554

Chenjingliang1 opened this issue Nov 7, 2023 · 2 comments
Labels
question Further information is requested

Comments

@Chenjingliang1
Copy link

Description

https://github.com/FlagAI-Open/FlagAI/blob/master/flagai/model/aquila2/modeling_aquila.py#L421
想问下函数参数这里写的是float的原因是啥，为什么不能是half的？
如果想用half的attention该怎么修改，尝试在这里改输入的数据类型，后面会有数据类型不匹配的报错。

Alternatives

No response
@Chenjingliang1 Chenjingliang1 added the question Further information is requested label Nov 7, 2023
@BAAI-OpenPlatform
Copy link
Collaborator

后面会在哪里报错呢

@Chenjingliang1
Copy link
Author

Chenjingliang1 commented Nov 9, 2023 •
edited
Loading

后面做matmul会报数据类型不一致的错

后来我改了脚本传入的deepspeed配置文件，里面配了fp16 enable，走了fp16的训练了。

还想问下脚本里 --fp16，这个是代表啥意思？

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
question Further information is requested
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
2 participants
@Chenjingliang1 @BAAI-OpenPlatform