NLP Neural Networks
Silas Liu — Feb 12, 2023
With computational resources evolving faster and faster, Artificial Intelligence is also evolving, presenting new and interesting structures and topologies.
We delve deep into the structures and applications of deep neural network, that evolved according to the needs and applications, and culminated in the state-of-art topologies nowadays, as the one used by ChatGPT.
Through development and application examples, we explain in a very simple way Encoder/Decoders, Autoencoders, LSTM (2014) and Transformers (2017). We showcase here several applications: image de-noising, anomaly detection, text generation, language translation and question answering.
The important results are shown here, but the entire code can be accessed through the button below. For the neural networks we used tensorflow, keras and transformers libraries.