Version 0.6.0

• Many major frontend features implemented.
• Structure & Advanced Controls panels for setting up the network.
• Data Preparation panel for uploading files and performing any necessary encoding on the data.
• Codec Class in the frontend responsible for encoding/decoding data with categorical columns, only numerical data should be passed to the backend.
• Training panel features many functionalities for customizing the training process (speed, epochs...) and an Error Chart displaying training & testing errors in real time.
• Testing panel shows the predictions and allows for custom tests.

Version 0.5.0

• First fully functioning release.
• Website is ready to use, data can be trained and tested.
• Some frontend features still need to be added.

Version 0.4.0

• Emscripten Complete Interface.
• Frontend can now fully control the Network in Wasm.
• D3.js for Network & Data Visualization.
• Interface is ready, frontend is all whats left.

Version 0.3.0

• Frontend using Hugo static generator.
• Full GitHub Actions pipeline ready to Test, Build, and Deploy.
• Emscripten wasm-compiler Integration.
• You can start writing WebAssembly!

Version 0.2.0

• First successful training.
• Comprehensive tests for all classes.
• Loss functions.
• XOR successful training example.
• No Web Assembly yet.
• No fruit classification dataset yet.

Version 0.1.0

• Core library is ready.
• All neural network components are implemented.
• No tests.
• No training data.
• No web assembly.