search

zjhuang22 / maskscoring_rcnn

Codes for paper "Mask Scoring R-CNN".
MIT License
1.9k stars 378 forks source link

loss函数为nan #116

Open xiaotiancai899 opened 2 years ago

xiaotiancai899 commented 2 years ago

开始训练模型后，所有的loss函数都是nan值，该怎么改呢？求助！！！！！！！！

248528e597dfd98255a86c3950dc673

wanghuogen commented 2 years ago

Have you solved this problem? I also met same problem when I trained with our own dataset.

kafei123456 commented 2 years ago

我改了base_LR=0.0002 和 defaults.py中的_C.SOLVER.IMA_PER_BATCH=2 以及 NUM_CLASS = 82。成功了

kafei123456 commented 2 years ago

你解决了这个问题吗？当我使用我们自己的数据集进行训练时，我也遇到了同样的问题。

我改了base_LR=0.0002 和defaults.py中的_C.SOLVER.IMA_PER_BATCH=2 以及NUM_CLASS = 82。成功了

zhaozhen2333 commented 2 years ago

This problem seems to be solved when using two or more GPUs