Coder Social home page Coder Social logo

Yi-6B 推理输出异常 about yi HOT 7 CLOSED

01-ai avatar 01-ai commented on August 26, 2024
Yi-6B 推理输出异常

from yi.

Comments (7)

Leymore avatar Leymore commented on August 26, 2024

是因为需求 torch==2.1.0 吗....

from yi.

cArlIcon avatar cArlIcon commented on August 26, 2024

请问环境中有安装flash attention吗?

是因为需求 torch==2.1.0 吗....

请问环境中有安装flash attention吗?如果有请升级flash attention到最新版本

from yi.

Leymore avatar Leymore commented on August 26, 2024

flash_attn 的版本是 2.0.0.post1,且不太方便升级。

请问 flash_attn 需要升级到 2.0.0 以上就行,or 必须要求最新版本?

from yi.

ZhaoFancy avatar ZhaoFancy commented on August 26, 2024

flash_attn 的版本是 2.0.0.post1,且不太方便升级。

请问 flash_attn 需要升级到 2.0.0 以上就行,or 必须要求最新版本?

目前我们只在最新版以及其它几个版本上测过,建议升级到目前的最新版后再测试下看看?

from yi.

cArlIcon avatar cArlIcon commented on August 26, 2024

flash_attn 的版本是 2.0.0.post1,且不太方便升级。

请问 flash_attn 需要升级到 2.0.0 以上就行,or 必须要求最新版本?

flash attention 在2.0.x的版本做inference会遇到点问题,升级到最新版结果应该就会正常。

from yi.

Leymore avatar Leymore commented on August 26, 2024

话说 2.0.x 具体会碰到什么问题?

我试了下我这个 2.0.0.post1 的版本,肉眼看输出结果已经正常了。问题解决前后的差异,应该仅在于 torch==1.13.1torch==2.1.0

from yi.

cArlIcon avatar cArlIcon commented on August 26, 2024

话说 2.0.x 具体会碰到什么问题?

我试了下我这个 2.0.0.post1 的版本,肉眼看输出结果已经正常了。问题解决前后的差异,应该仅在于 torch==1.13.1torch==2.1.0

2.0.x应该会导致你上面提到的状况,建议升级到2.3.x

from yi.

Related Issues (20)

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.