Comments (2)
Sorry, I didn't properly read that you already have set up data loading
I did have to make some tweaks in my fork in order to load RGB images, and also to change the number of logits on the output, specifically the num_outputs
argument in the output FC layer in vectorCapsNet.py. I also wanted to experiment with setting the learning_rate for Adam Optimizer so I made it a hyper parameter, and found decent results with 0.0001 (default being learning_rate=0.001
) which you can of course just manually set
I've noted the num_outputs issue here, #10
The current master of my fork isworking for the binary classification of gender on the RGB, 299x299px LFW Faces Dataset up to about 95% accuracy, but decoded reconstruction isn't great, I might need to investigate some larger Fully Connected layers in the decoding stage.
If you're still having issues could you perhaps post a picture of your TensorBoard graphs?
from capslayer.
You will need to change the parameters on line 124 of what image size to expect, but then you will need to ensure your custom data is being loaded in; have you properly set up your custom dataset loading in capslayer/utils.py
?
main.py will load the dataset defined by cfg.dataset and call the appropriate dataset handler in utils.py, for example load_mnist
from capslayer.
Related Issues (20)
- ValueError when using cl.layers.conv2d HOT 1
- Reproducing results from "Matrix Capsules with EM Routing" HOT 2
- EM Capsule Dense Layer Routing HOT 4
- Support for MSCOCO dataset HOT 1
- Spelling Error
- CapsLayer Neural Network Example
- Question for E_step in EM Routing
- Prediction problem HOT 3
- ModuleNotFoundError: No module named 'capslayer.data.datasets.stanford_drone' HOT 2
- cifar10 dataset "Maximum allowed size exceeded" Error HOT 2
- tensorflow2.0 how to write config.py? HOT 1
- i have issue with cifar10
- OutOfRangeError (see above for traceback): End of sequence
- Hi, You can show an example application with CONV3D
- Routing by agreement with Transformer-based for NMT
- In the provided MNIST example not all capsules are seeing activation probability
- fashion mnist support
- Multi-label classificatoin
- Performance issues in capslayer/data/datasets/cifar10/reader.py
- Performance issues in /capslayer/data/datasets (by P3) HOT 1
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 capslayer.