-
Notifications
You must be signed in to change notification settings - Fork 345
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
infer的时候,如果问训练集问题,模型响应异常 #27
Comments
修改了torch/nn/modules/linear.py的114行,把self.weight(即attention的linear权重)所在device设为合适的cuda: |
我自己也用自己的数据训练了,也是这样,不知道为啥?但是alpaca_chinese_dataset训出来的模型没问题,是数据问题吗? |
而且,反复打印的输入参数显示,输入文本越来越长;使用训练epoch更久的模型,基本都会响应异常(这点也对应了我先前没提到的另一个奇怪现象,训练过了4个epoch之后,train_loss和eval_loss都开始变很大)。
我也是用自己数据训练,和LoraConfig的target_modules设置有关系吗? |
使用最新代码微调之后,infer的时候遇到了以下bug,不知道是不是lora调整的层加载在cpu?还是别的原因
The text was updated successfully, but these errors were encountered: