-
Notifications
You must be signed in to change notification settings - Fork 418
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
微调完成后推理加载问题 #61
Comments
请改成
|
如果数据多的话可以考虑增加训练的参数,目前的训练脚本只训练了2层lora,可以看到 https://github.com/THUDM/VisualGLM-6B/blob/main/finetune_visualglm.py#L18 的layer_range参数,目前只finetune了[0,14]两层的lora。 |
实在是太感谢你了,在微调的过程中给予我这么多帮助,我看到这行代码了 我知道在SD里lora层数是26层 |
|
layer_range接受一个list作为参数,代表要加lora的层编号,比如我们提供的脚本里 |
感谢指导,visualglm是否支持chatglm一样的多轮对话训练?
|
支持,可以参考 |
python3 cli_demo.py --from_pretrained '/root/ai/VisualGLM-6B/checkpoints/finetune-visualglm-6b-05-25-15-02/5000/mp_rank_00_model_states.pt' --prompt_zh 请给这张图片写一段广告语?
根据官方的命令修改了一下,报错信息
The text was updated successfully, but these errors were encountered: