Even

icon indicating a website link https://github.com/CaptainEven icon indicating an email [email protected]

icon location Wuhan China icon location Algorithm engineer

Results 105 comments of Even

gen_labels

ok

gen_labels

应该不行，heatmap数量不一致

gen_labels

是的

No param id.0.weight.If you see this, your model does not fully load the pre-trained weight. Please make sure you have correctly specified --arch xxx or set the correct --num_classes for your own dataset.

@Ckq-Sugar 这不是报错，是warning，忽略就行了，能不能出来 类似这样的 ![image](https://user-images.githubusercontent.com/9303747/86879917-80110480-c11e-11ea-9945-43905a2710d7.png) 训练日志信息？

No param id.0.weight.If you see this, your model does not fully load the pre-trained weight. Please make sure you have correctly specified --arch xxx or set the correct --num_classes for your own dataset.

@Ckq-Sugar 训练成功运行，会实时打印类似这样的信息: mot/default |########################## | train: [1][15176/18030]|Tot: 5:50:16 |ETA: 1:05:55 |loss 0.7844 |hm_loss 0.4438 |wh_loss 1.3069 |off_loss 0.2037 |Data 0.002smot/default |########################## | train: [1][15177/18030]|Tot: 5:50:18 |ETA: 1:05:50 |loss 0.7844 |hm_loss...

No param id.0.weight.If you see this, your model does not fully load the pre-trained weight. Please make sure you have correctly specified --arch xxx or set the correct --num_classes for your own dataset.

上面的排版有点问题，总之会打印epoch, batch, loss等信息

pytorch vision

@wangshuo6699 @qq542288564 pytorch0.3.1吧，改成pytorch1.0及以上很容易，去掉Variable即可。

多GPU训练问题

@ZacharyZgw 用的哪个版本的PyTorch?

Is this method compatible with TorchScript?

I did not tried that, you could have a try, if succeed, there can be a version merge, looking forward for JIT accelerating!

[Layer support] How to implement a grouped softmax layer

@WongKinYiu Thank you for the answer!