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

使用 8 x 4090 微调 llama3-1-8B 后,询问无输出反馈。 #241

Open
Sakkana opened this issue Aug 15, 2024 · 5 comments
Open

使用 8 x 4090 微调 llama3-1-8B 后,询问无输出反馈。 #241

Sakkana opened this issue Aug 15, 2024 · 5 comments

Comments

@Sakkana
Copy link

Sakkana commented Aug 15, 2024

如图,跑了一遍仓库里的代码。
微调结束后,使用 checkpoint-699,输入 prompt 后大模型没有给任何输出。
求好心人解答这是怎么回事

截屏2024-08-16 06 36 10
@Sakkana
Copy link
Author

Sakkana commented Aug 15, 2024

@KMnO4-zx 大佬😭

@Sakkana
Copy link
Author

Sakkana commented Aug 15, 2024

有重启 notebook。
response 长度为 0,如下图。

截屏2024-08-16 06 45 47

@LinB203
Copy link

LinB203 commented Aug 20, 2024

same question.

@Sakkana
Copy link
Author

Sakkana commented Aug 20, 2024

update:

base模型和lora权重合并后保存为新模型(merge_and_unload、save_pretrained)后,使用vllm可以完成部署推理。

#教程中的推理方式无输出。(仅限llama3.1,Qwen2-7b、bilibili-index、DeepSeek等正常)

@Tsumugii24
Copy link
Contributor

Tsumugii24 commented Sep 19, 2024

update:

base模型和lora权重合并后保存为新模型(merge_and_unload、save_pretrained)后,使用vllm可以完成部署推理。

#教程中的推理方式无输出。(仅限llama3.1,Qwen2-7b、bilibili-index、DeepSeek等正常)

可能是版本的问题,最近更新了 requirments ,可以按照新的 package version 来试一下,我个人在目前的环境配置中没有遇到问题

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

3 participants