yxizhong
yxizhong
Dear authors, Thank you for your work on Robust Adaptation of Foundation Models with Black-Box Visual Prompting, in which the method of PCA extraction of statistical information in BlackVIP-SE is...
您好!非常感谢您的开源! 我现在需要做一个非文档的特定数据dewarping工作,如果直接使用你提供的docres.pkl,模型输出是正常的,但是效果不是很好。 我现有的数据是原图image和图像中物体的位置掩码mask,我通过image和mask计算了类似doc3d的bm(通过warp验证了bm应该是正确的),但是使用这个数据集直接训练/微调docres.pkl后,模型的输出都变成全黑了?请问这是什么情况呢? 期待您的回复!
Thank you for the open source! I am trying to train a model using the framework and the SFT phase seems to be working, but the RL phase makes a...