Fixing a LSTM Text Generator: Update #4

To be honest, I saw this one coming. Okay, maybe I didn’t expect this to be my biggest problem. I had only heard of overfitting in relation to machine learning classifiers giving the wrong labels to images or something. I didn’t expect that my headline generators would be suffering from overfitting because I didn’t realize […]

LGBTQ Content Filtering: Update #3

So I have been continuing to look at the Facebook ad library and policies, and I realized I have made a mistake. When searching through the Facebook ad library, searching with keywords as I have been doing (like “lgbt”, “lgbt”,”gay”, “lesbian”, etc) ONLY shows ads that have the “social issues, elections, or politics” classification. In order […]

Generating Text With Neural Networks: Update #3

Now that I have a minimal understanding of what neural networks are and how they work, I figured it was time to embark on the journey of developing text generation networks. Using a few online tutorials, particularly this one , I was able to create a few small text generation networks. This tutorial was focused on creating […]