Yuetian Wang

确实是这样

Lol, really wish it was possible to open up the ability to use text prompts . A two-stage approach like Grounded-Segment-Anything is neither useful nor elegant.😣

Translate "How to set the interval parameter" to English -> [PR#959](https://github.com/open-mmlab/mmengine/pull/959)

Because of the irregular naming of branch, PR #959 has been closed and all updates from it have been moved to PR #964.

> 作者您好，请问论文中的实验结果可以用您在论文中设置的超参数复现出来吗，我这边使用了您在论文中提到的超参数设置（近邻簇为1500中取1300个，并进行90%的节点抽样，use_Sim = True，阈值设置为0，knn设置为80），训练到第一个epoch的3899batch，训练loss降为0.02左右，但在测试集上的pairwise ave_pre很低，不明白为什么？谢谢~ 遇到了同样的问题