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

[Docs] Fine_Tune #859

Open
Newbie-pz opened this issue Jan 15, 2025 · 1 comment
Open

[Docs] Fine_Tune #859

Newbie-pz opened this issue Jan 15, 2025 · 1 comment

Comments

@Newbie-pz
Copy link

📚 The doc issue

您好
我想问一下这个微调代码应该如何执行 是下载好intervl2.5_mpo文件然后用pycharm的terminal直接执行吗GPUS=1 PER_DEVICE_BATCH_SIZE=1 sh shell/internvl2.5_mpo/preference_optimization
/internvl2_5_1b_qwen2_5_0_5b_dynamic_res_mpo.shGPUS=1 PER_DEVICE_BATCH_SIZE=1 sh
还有就是我要做一个医学影响方面的识别对话模型 用internvl2.5作为基座模型 我是不是应该从internvl_chat/internvl/model/internvl_chat进行架构的修改

Suggest a potential alternative/fix

No response

@yuecao0119
Copy link
Collaborator

你好,

  1. 微调步骤可以参考我们的教程,MPO和InternVL2.0/2.5步骤基本一致。另外我们的shell脚本是基于slurm进行计算节点管理,你需要根据自己集群管理方式进行适当修改。
  2. 如果你只是需要新的领域的对话能力,只需要加入对应数据集进行微调即可,如果是对模型框架进行修改,你是对的,在internvl_chat/internvl/model中修改模型。

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