Hi all, I recently published a framework on GitHub for creating neural networks, including feed forward, recurrent (RNN/LSTM) and transformer models. It's built from scratch and is meant to be a simpler alternative to Keras for sequential data/time series forecasting. It's really fast too. Is this something that could be of interest to practitioners? Looking for feedback! Thanks
adetrick•1h ago