alekscipri / deepmergedomainadaptation Goto Github PK
View Code? Open in Web Editor NEWFisher deep domain adaptation for astronomy - galaxy mergers
License: BSD 2-Clause "Simplified" License
Fisher deep domain adaptation for astronomy - galaxy mergers
License: BSD 2-Clause "Simplified" License
Compare simple training VS training+transfer loss (MMD/Coral or ADA) VS training+full domain adaptation
Might be good to put the requirements.txt file back - or is it just PyTorch, basically? Is the code using any "niche" libraries?
Also, it might be neat to put a copy of a "working" Colab notebook here (just save as a ipynb and stick it in a "demos" folder).
Add old DeepMerge CNN to our code.
Which ResNet do we want to use?
How do we want to train it? Frozen layers or not? Using transfer learning from imagenet weights or not?
Rewrite code to use our numpy data.
Final step - add path to the dataset as an argument for training
Change early stopping to monitor validation loss, wince this is best for preventing overfitting.
Figure out how training loss and accuracy is currently calculated. Add validation training and accuracy and save them in the log file so we can monitor them using tensorboard.
Normalize based on mean and std of training set only!
Remove all things we don't need. VGGs and Alex net for example.
Also all arguments we don't use.
Perhaps also everything regarding loading images and numpy arrays?
A declarative, efficient, and flexible JavaScript library for building user interfaces.
๐ Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. ๐๐๐
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google โค๏ธ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.