Scalsol / mega.pytorch

Memory Enhanced Global-Local Aggregation for Video Object Detection, CVPR2020
Other
566 stars 115 forks source link

我用官方给的配置文件和数据集训练模型得到结果并没有论文中的好,我的是78.1%,这是什么原因呢? #96

Open asmallcat opened 3 years ago

cecilia930426 commented 3 years ago

Same here

asmallcat commented 3 years ago

同样在这里

需要同比例的调整lr和step,可参考https://github.com/facebookresearch/Detectron/blob/master/configs/getting_started/tutorial_1gpu_e2e_faster_rcnn_R-50-FPN.yaml#L14-L30

SouperO commented 2 years ago

遇到同样的问题,使用1个GPU,同比例调整了lr和step,包括相同的配置文件和数据集训练,只能达到78.3%,请问卡的数量对结果的影响会有这么大吗?