-
代码数据读取正常,网络框架正常,但整体拼接在一起时使用optimizer.minimize(avg_cost) 报错,似乎是说无法反向传播
训练部分代码如下:
```
from dataloader import *
from IresNet import IresNet
from loss import model_loss
if __name__ == '__main__':
…
-
There is no tutorial about two or more model inputs in doc, is it supported?
-
When I try to run the predict file using the pictures I took, I get very poor prediction results. Is there something wrong with me?
![QQ截图20200723223836](https://user-images.githubusercontent.com…
-
-
I train your network without deformable conv and refinement.
The train err on sceneflow is high, about 2.0 epe.
Does it normal?
-
When I tried to train on the sceneflow data set, I encountered this problem, but I did not have this problem in previous training. How should I solve it?
`
Number of model parameters: 3174132
590…
-
Thank you for your excellent work! When I check the number of epochs in the AcfNet configs for training on SceneFlow dataset, I found total_epochs is 20, but the checkpoint is set to epoch_10.pth. May…
-
**commond** : "python predict.py --data_dir demo --pretrained_aanet pretrained/aanet+_sceneflow-d3e13ef0.pth --feature_type ganet --feature_pyramid --refinement_type hourglass --no_intermediate_superv…
-
my environment is pytorch 1.1 and cuda 10.0 .
In compiling, there is no error occuring.
But when running ganet_deep.py , sync bn is wrong. Then i replace sync BN by pytorch common BN.
At last, eve…
-
my test code of time costs is as following:
```python
def test(imgL, imgR):
model.eval()
with torch.no_grad():
imgL = Variable(torch.FloatTensor(imgL))
imgR = Variable(to…