Scalsol mega.pytorch issues

Scalsol / mega.pytorch

Memory Enhanced Global-Local Aggregation for Video Object Detection, CVPR2020

Other

570 stars 116 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Fixed deprecation issues and made compatible with newest pytorch and CUDA versions

#125 AmetistDrake opened 2 weeks ago
0
求助！训练自己的数据集出现loss为nan

#124 xixioo123 opened 1 month ago
2
How can it support cuda 11.x?

#123 lively2018 opened 2 months ago
0
ImportError: DLL load failed: 找不到指定的程序。

#122 fengjingchehu opened 1 year ago
0
Dataset unavailable

#121 LucaRo29 opened 1 year ago
0
INSTALL.md Problem

#120 ComeOnComeOnTurnYourRadioOn opened 2 years ago
1
Bug Question about validation while training

#119 CyberPegasus opened 2 years ago
0
求助大佬们，未修改代码，直接跑的，结果全是0

#118 hehew321 opened 2 years ago
4
apex version not match pytorch version 1.3 or lower

#117 LilyDaytoy opened 2 years ago
3
failed running python setup.py build develop for mega.pytorch

#116 LilyDaytoy opened 2 years ago
7
Can I change the backbone of MEGA? I try to use SWIN,VIT,VIG as the backbone，but I got 0.01% Map.

#115 kangkang189 opened 2 years ago
0
windows mmdeploy SDK构建失败

#114 HuangLonghao closed 2 years ago
0
when i use 1 gpu training， how to change the iters、step、lr thank you

#113 caofuxiao321 opened 2 years ago
0
why use the DET datasets?

#112 yuyifan1991 opened 2 years ago
0
i test the dff(backbone is resnet101),but the AP50 is 66.58，is much lower than you publish，who can tell why?

#111 liubinbinbin opened 2 years ago
0
How to train with ONLY VID？

#110 youzi260 opened 2 years ago
0
TypeError: __call__() takes 2 positional arguments but 3 were given

#109 lemyanh2612 closed 2 years ago
1
there was a probblem using my own datasets:RuntimeError: Sizes of tensors must match except in dimension 3. Got 425 and 423 (The offending index is 0)

#108 LuoXiang-LX opened 2 years ago
3
I have encountered a problem when using my own data set for training and evaluation.

#107 XiaoSue opened 3 years ago
1
A severe BUG in dataloader (mega_core/data/datasets/vid_mega.py)

#106 Dawn-LX opened 3 years ago
0
How to aggregate the results of local aggregation and LGlobal aggregation?

#105 longmalongma opened 3 years ago
0
How to plot the loss/P-R curve

#104 MrunaliniN opened 3 years ago
0
About reproducing FGFA results

#103 HzZHoO closed 3 years ago
0
pic should be Tensor or ndarray. Got <class 'NoneType'>.

#102 Flyingdog-Huang opened 3 years ago
0
RuntimeError: CUDA out of memory.

#101 Flyingdog-Huang closed 3 years ago
1
请问下面两个命令哪里有问题，还是本身就不支持？（我的想法是：用MEGA-R101的模型去推理我自己的数据）

#100 Flyingdog-Huang closed 3 years ago
1
FileNotFoundError: [Errno 2] No such file or directory: 'R_101.pth'

#99 Flyingdog-Huang closed 3 years ago
0
How to train on ICDAR2013 dataset?

#98 cyborg1995 opened 3 years ago
0
在单GPU下进行消融实验效果比完整的MEGA效果好

#97 asmallcat opened 3 years ago
3
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 asmallcat opened 3 years ago
3
这个跑不起来呀，又谁跑起来了吗

#95 asmallcat opened 3 years ago
1
fix typo

#94 jokober closed 3 years ago
0
Results of single frame baseline

#93 DavideA opened 3 years ago
0
产生许多权重文件，如何判断那个是最好的

#92 joe660 opened 3 years ago
0
MEGA QQ交流群

#91 joe660 opened 3 years ago
1
What is the significance of multi relation modules？

#90 zhanghaoo opened 3 years ago
0
The architecture of MEGA？

#89 zhanghaoo opened 3 years ago
0
请问训练FGFA是用什么命令？

#88 CHP-2020 opened 3 years ago
1
VID_val_frames.txt和VID_val_videos.txt

#87 joe660 closed 3 years ago
0
你好作者关于复现出现的问题

#86 joe660 closed 3 years ago
0
有

#85 joe660 closed 3 years ago
0
About mAP metrics

#84 ruiyuanlu opened 3 years ago
0
How exactly 'flownet.ckpt' was trained?

#83 pyCERN opened 3 years ago
0
How do you visualize the output of FlowNet in DFF

#82 gxpzy opened 3 years ago
2
Should I use the 4-step training method to train GeneralizedRCNN

#81 gxpzy closed 3 years ago
0
测试代码里图片可能加载错了

#80 jianguolalala opened 3 years ago
0
How is single frame baseline trained?

#79 KeepGoing2019HaHa closed 3 years ago
1
how to get flownet.ckpt?

#78 jhxu-org opened 3 years ago
0
关于其中的FGFA，很可能有错误，请看下面

#77 fengyue20 opened 3 years ago
6
Performance issue

#76 lxtGH closed 3 years ago
7