brianakl Goto Github PK

bootstrapped_model_up_scaling

Experimentation with creating a way to transfer knowledge from a smaller transformer model to a larger one

brian-naklycky-resume

My resume

cifar_cnn

Using CNNs to classify the CIFAR-10 and CIFAR-100 datasets

distilbert_learning_to_skip_text

This is a model used to compare the significant speed and accuracy increases in the use of transformers in comparison to pre-transformer architectures. The model it was compared to was Learning to Skim Text Wei Lu et. al. https://aclanthology.org/P17-1172/

mnist_nn_experiment

Experimenting with the impact of noise on NN training

nlp_final_paper

Our goal is to find Dataset Artifacts and hard examples using Dataset Cartography and focus our models learning on those examples. We use the SNLI dataset from to test techniques discussed in on finding data artifacts as well as using dataset cartography to improve our performance

nvim_config

My personal nvim config, primarily use python and c/c++

sudoku_solver

A sudoku solver using dancing links algorithm

traveling_salesman