App wrapped version of a previous project which can be found here where a neural network generates realistic nonsense adjectives that the user can then define into a 'dictionary' entry.
Uses multithreading to enable the neural network to continue generating words in the background, while the user interacts with the app on a different thread.
The app will show the user words that have been generated live, just in the time since the app was launched, which means there is a slight initial delay and potential for further small delays depending on the speed at which the user uses the app or how often they skip words. However, the app will automatically redirect the user to a loading screen or the main functionality as needed, if the words run out or are generated.
Made using Kivy/KivyMD, Python and Keras/Tensorflow.
The model in this project was also trained on a collection of real adjectives from this dataset. Thank you to the creator of the dataset!