Comments (2)
Sorry that I did not integrate Tensorboard in this code. However, I write a very simple logger to save the training progress. After training, you should find the log.txt
in the directory when you save your model (e.g., checkpoints/cifar10/alexnet/log.txt
). The log file should look like this
Learning Rate Train Loss Valid Loss Train Acc. Valid Acc.
0.100000 1.451691 1.147165 47.110000 59.760000
0.100000 0.914918 0.928320 67.084000 68.180000
0.100000 0.719559 0.758013 74.574000 74.240000
0.100000 0.604210 0.751205 79.030000 75.660000
0.100000 0.538204 0.562017 81.564000 80.490000
0.100000 0.493776 0.669323 82.940000 77.790000
0.100000 0.465779 0.521751 83.890000 82.960000
0.100000 0.437683 0.601297 84.980000 79.930000
Then you can use the following script to visualize and compare the training process of the different models. For example, comparing the Valid Acc.
for densenet-100 and densenet-190,
from utils import *
import matplotlib.pyplot as plt
paths = {
'densenet100':'/home/wyang/code/pytorch-classification/checkpoints/cifar10/densenet-bc-100-12/log.txt',
'densenet190':'/home/wyang/code/pytorch-classification/checkpoints/cifar10/densenet-bc-L190-k40/log.txt',
}
field = ['Valid Acc.']
monitor = LoggerMonitor(paths)
monitor.plot(names=field)
savefig('test.eps')
plt.show()
If you are willing to integrate Tensorboard into this code, feel free to create a PR. Thank you.
from pytorch-classification.
Thanks @bearpaw
FYI, tensorboard_logger
from pytorch-classification.
Related Issues (20)
- Missing densenet-bc-100-12 weights for cifar100 on OneDrive
- running with a newer pytorch version HOT 1
- For vgg16, there are three classifier layers in the provided checkpoint but only one in the model HOT 2
- Error loading pretrained model weights HOT 1
- The pretrain cifar10 resnet110 indeed is resnet164 (BottleNeck) HOT 1
- PreResNet-110 on Cifar100, Top1 error rate is 26.47 rather than 23.65. HOT 3
- about how to inference HOT 2
- 'ProgressBar' object has no attribute 'elapsed_td'?
- The parameters count is different from torchsion resnet.
- CIFAR-10 does not have resnet18
- draw the accuracy curce such as ./utils/images?
- How long does ImageNet take to train?
- the result of resnet18 with imagenet is Test Acc: 0.09(I loaded your pretrained model)
- Inconsistent ResNeXt-29 (16x64) trained model on CIFAR10
- tensor(89.0110, device='cuda:0')
- Problems with the checkpoints of the PreResNet110
- The results of ResNeXt-50 (32x4d) on ImageNet
- Checkpoints are unaccessible HOT 1
- RuntimeError: view size is not compatible with input tensor's size and stride (at least one dimension spans across two contiguous subspaces). Use .reshape(...) instead. HOT 1
- why depth of resnet-110 are 164?
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 pytorch-classification.