-
Notifications
You must be signed in to change notification settings - Fork 2.5k
Open
Labels
bugSomething isn't workingSomething isn't working
Description
出bug的具体模型
Internalm2-chat-7B
出bug的具体模型教程
使用xtuner进行微调,如transformers的等等库版本不协调
教程负责人
天机教程贡献者
Bug描述
要不出现ImportError: cannot import name 'StaticCache' from 'transformers.cache_utils' (/home/qlunlp/anaconda3/envs/xtuner/lib/python3.10/site-packages/transformers/cache_utils.py)
要不出现 Cache only has 0 layers, attempted to access layer with index 0
很多版本错误,导致最后的xtuner train ./internlm2_chat_7b_qlora_oasst1_e3_copy.py --deepspeed deepspeed_zero根本跑不起来。
复现步骤
按照教程中建库部分自己比着敲一下就知道了。
期望行为
xtuner train ./internlm2_chat_7b_qlora_oasst1_e3_copy.py --deepspeed deepspeed_zero2
能跑
环境信息
python 3.10
linux
其他信息
无
确认事项 / Verification
- 此问题未在过往Issue中被报告过 / This issue hasn't been reported before
Metadata
Metadata
Assignees
Labels
bugSomething isn't workingSomething isn't working