Masters

Architecture Stack      |      README      |      Model Research

Model Research

Research into the different models that came before and which areas to keep and what could be improved. Privacy BERT-LSTM and Polsis were good models, however their lack of tokenization caused a bottleneck in the amount of data the model could process and keep track of. The addition of the LSTM transformer allowed for short-term memory to be kept for a longer period of time, but the word embedding was not made for longer pieces of text.

This new model purposed is called Prert-CNM (Pr-ivacy B-ERT Contextual Neural Memory) increases the number of specialised parameters (increased embedding layer) and tokens (100 Input w/ 512 limit to 4096), swaps the general purpose BERT-LSTM model with the specialised mode DeBERTa v3 (with help from several other models working as helper agents) to process larger docs and hold onto more memory.

Chapters


Prert-CNM

Table of Contents

Memory

Context

Weightings

Cluster Control

LSTM Layer

Dataset

ISO Controls

Model Architecture

Hierarchical Transformer (Uses)

Output

Returns back