Documented and Unit Tested educational Deep Learning framework with Autograd from scratch.
New operation, bug fixes.
Fixed an error in the Dropout layer. Now it should work normally.
Added Xavier initialization feature to .randn() and .randn_like() methods, for better weight initialization. Also, added workflow to automatically remove previous deployment files.
Added Xavier initialization feature to .randn() and .randn_like() methods, for better weight initialization.
Fixed deployment workflow. Improved documentation.
Fixed deployment workflow. Added better documentation.
Simplified code and added more documentation.
Fixed workflows
Added workflows and documentation improvements.
This is a complete version, with documentation and unit tests. Further versions will contain more operations, and GPU support. PyPi link: https://pypi.org/project/neuralforge/