Project-HAMi / HAMi

Heterogeneous AI Computing Virtualization Middleware
http://project-hami.io/
Apache License 2.0
957 stars 197 forks source link

请问支持多机吗 #565

Open HeatherLiuzh opened 1 month ago

HeatherLiuzh commented 1 month ago

Please provide an in-depth description of the question you have:

What do you think about this question?:

Environment:

archlitchi commented 1 month ago

请问具体的需求是?

HeatherLiuzh commented 1 month ago

想做分布式训练,多级多卡这种

limengxuan @.***> 于2024年10月21日周一 17:24写道:

请问具体的需求是?

— Reply to this email directly, view it on GitHub https://github.com/Project-HAMi/HAMi/issues/565#issuecomment-2426108505, or unsubscribe https://github.com/notifications/unsubscribe-auth/BMCSRY6DJKBOZCXKM2R4VDTZ4TB5ZAVCNFSM6AAAAABQJV2IXSVHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHMZDIMRWGEYDQNJQGU . You are receiving this because you authored the thread.Message ID: @.***>

taikai-zz commented 3 weeks ago

我的需求主要是推理用,多机多卡的问题,一台机器的显存无法装载一个大模型,需要多机多卡,请问支持吗?@archlitchi

yule-sun commented 2 weeks ago

同問,支持分佈式訓練,跨主機分配顯卡資源不?

leeyuejun commented 3 days ago

分布式训练,个人感觉还是应该用原生的卡, 既然一个卡都不够用, 就没必要再切割了.