Comments (7)
You can calculate the softmax loss once and generate a mask for postive instances and hard negtive instances. Then you can multiply them together to generate the valuable loss.
from pytorchssd.
@taneslle , HI, have you figured it out why?
from pytorchssd.
Because we want to get the softmax loss for each element. However, in the early version of Pytorch, this function is not supported and only the sum or average loss of all the elements is produced.
from pytorchssd.
@lzx1413 Thank you so much for your replying! Now, we can control the 'reduce=False" to get the element -wise loss. So, do you mean we can just use the F.cross_entropy
twice now? One for the hard negative mining, one for the calculation of the final conf_loss ?
from pytorchssd.
Yes. Another issue is that if we can do it by using the F.cross_entropy twice. Will it change the input tensor's gradient because the forward pass of the same input is tracked. Thank you for your help again!
from pytorchssd.
If you don't add the first softmax's loss to the final total loss, it will not affect the network.
from pytorchssd.
Thank you.
from pytorchssd.
Related Issues (20)
- IndexError: too many indices for array HOT 1
- fssd divergent HOT 1
- No module named 'yacs'
- 'RFBNet' object has no attribute 'module'
- TypeError: Expected Ptr<cv::UMat> for argument 'src' HOT 1
- Unexpected key(s) in state_dict: L2Norm.weight HOT 2
- The mAP of SSD-Resnet is lower than SSD-VGG16. HOT 1
- 想问下大家,在训练的过程中又碰到AP为0的情况吗,我是用的自己的数据集训练的。num_class啥的都改过了,实在不知道该咋办了
- from .nms.cpu_nms import cpu_nms, cpu_soft_nms ImportError: /home/gabriel/PytorchSSD/utils/nms/cpu_nms.so: undefined symbol: _Py_ZeroStruct
- all loss is nan
- Hi, I need your help ! RefineDet320 voc 2007 only get mAP=0.76, lower than SSD300? HOT 4
- No module named 'utils.nms.cpu_nms'
- Where is L2Norm???
- What is the RFB_E model in models/ folder?
- Why Refinedet only support 320 input dim? HOT 1
- batch_size=16 or batch_size=32 ? HOT 1
- Where can I download other pre-trained mobilenet models?
- RuntimeError: copy_if failed to synchronize: device-side assert triggered
- File "train_test.py", line 127, in <module> net = build_net(size=300, num_classes=7) File "/home/syh-hwj/daima/PytorchSSD/models/FSSD_vgg.py", line 224, in build_net return FSSD(*feature_transform_module(vgg(vgg_base[str(size)], 3), add_extras(extras[str(size)], 1024), size=300), File "/home/syh-hwj/daima/PytorchSSD/models/base_models.py", line 14, in vgg conv2d = nn.Conv2d(in_channels, v, kernel_size=3, padding=1) File "/home/syh-hwj/anaconda3/lib/python3.6/site-packages/torch/nn/modules/conv.py", line 297, in __init__ False, _pair(0), groups, bias) File "/home/syh-hwj/anaconda3/lib/python3.6/site-packages/torch/nn/modules/conv.py", line 33, in __init__ out_channels, in_channels // groups, *kernel_size)) RuntimeError: CUDA error: out of memory
- 每次到测试阶段都报错
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from pytorchssd.