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

models/Qwen/06-Qwen-7B-chat 全量微调 中deepspeed使用的是什么版本呢? #308

Open
1259010439 opened this issue Dec 6, 2024 · 4 comments

Comments

@1259010439
Copy link

我在进行 DS_BUILD_OPS=1 pip install deepspeed 会有如下报错
ModuleNotFoundError: No module named 'dskernels'

@KMnO4-zx
Copy link
Contributor

KMnO4-zx commented Dec 6, 2024

那就撞一下这个包呗

@1259010439
Copy link
Author

嗯嗯 这个问题已经解决了 现在是会有这个错误 我现在想训练的模型是 Qwen2.5-Coder-7B-Instruct 是这个的原因么?
[rank1]: File "/root/miniconda3/envs/deepspeed/lib/python3.10/site-packages/transformers/models/auto/configuration_auto.py", line 761, in getitem
[rank1]: raise KeyError(key)
[rank1]: KeyError: 'qwen2'

@KMnO4-zx
Copy link
Contributor

KMnO4-zx commented Dec 9, 2024

升级transformers版本就行了,这个报错是因为没有找到qwen2这个模型

@1259010439
Copy link
Author

还想再请教一下训练后的checkpoint 可以 使用vllm的 vllm serve Qwen/Qwen2.5-Coder-7B-Instruct --dtype=half --tensor-parallel-size 4 --download-dir Qwen/v1 --enable-lora --lora-modules lora_name=checkpoint-xx 这个方式直接加入到启动的server服务中么? 怎么能过验证一致这样的启动方式是否已经是加入了这些checkpoint呢

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