Closed zeeshanalipanhwar closed 3 years ago
Could any configurations there not have been set rightly?
I also have the same problem. When training the CondInst, the loss is always unstable. And after 90k training of R_50_1x, the loss is about 2.0± and can not detect any targets in COCO dataset.
One possible reason could be the way we are loading the data. Not sure where I am going wrong. Yet.
Thanks for you reply. Did you solve this problem? When training coco, my IMS_PER_BATCH=1 because of the limit of the memory, which causes the loss is very unstable. And now I try to increase the batch size by lower the MAX_PROPOSAL from 500 to 200 and clip the images. Then use two gpus to train the model together. How about your suggestions?
------------------ 原始邮件 ------------------ 发件人: "Zeeshan Ali"<notifications@github.com>; 发送时间: 2021年3月7日(星期天) 下午3:09 收件人: "aim-uofa/AdelaiDet"<AdelaiDet@noreply.github.com>; 抄送: "陈振乾"<693497091@qq.com>; "Comment"<comment@noreply.github.com>; 主题: Re: [aim-uofa/AdelaiDet] CondInst gives BBox AP:15 and Segm AP:0.0063 on Evaluation Set after being trained for 3K epochs! (#289)
One possible reason could be the way we are loading the data. Not sure where I am going wrong. Yet.
— You are receiving this because you commented. Reply to this email directly, view it on GitHub, or unsubscribe.
I did not solve the problem, and have left it there since two months. :)
Hi, I trained the CondInst on a custom dataset for 3000 epochs.
Training losses:
Validation results:
The model does not predict any BBox or Segm Mask for any inference sample I give to it. What could have gone wrong?
Selective logs for reference: