pigworms

Results 5 comments of pigworms

Excuse me, have you solved the problem that loss drops to 8.9 and then rises in the opposite direction? Is it caused by apex mixed precision training?

> 会不会是loss函数的问题呀 这个代码你还在关注吗,我的loss从开始就是16 降不下去

> > > 对,以前代码比较乱所以我重构改版了,新的dataloder用的是ct图像或者MR图像做其他项目的检测,重构后代码后没有直接跑luna16,重构前的v2版本跑过luna. > > > > > > 你好作者,请问您是否还能提供重构前的v2版本跑过luna跑过luna的代码。我调了你给的重构训练代码啥的,始终不太对。我的邮箱是 [[email protected]](mailto:[email protected])。 > > 你好,请问你的重构训练代码做好了吗,我最近也在研究这篇代码,但是现在卡在了数据集载入这里。如果您看到不知道可不可以分享给我,谢谢大佬!我的邮箱是 [[email protected]](mailto:[email protected])。 你好 请问你这个问题解决了吗,我也在尝试复现代码,卡在了数据处理这一部分

Hi, I'm a little confused about how to use FROCeval.py Can you show me what paths to fill in annotations_filename, annotations_excluded_filename and seriesuids_filename?

> Who can share the dataset after pre-processing, 2d or 3d is fine,thanks Do you have the preprocessed data? Can you send me a copy? Thank you!