jsbaan Goto Github PK

attentionexplanation

attn_vis

Interactive in-browser attention visualizer tool for recurrent networks

calibration-on-disagreement-data

Code accompanying the EMNLP 2022 paper "Stop Measuring Calibration When Humans Disagree" in which we show problems with popular calibration metrics like ECE in settings where more than one answer is acceptable, and argue for several metrics that take into account the full human judgement distribution.

computervision1

Implements image classification using handcrafted SIFT features as well as CNN extracted features for the course Computer Vision 1

cross-lingual-nmt-embeddings

This repo forms a refactor of the code for a project that focussed on Neural Machine Translation

datastruct-project

Java-based implementation of the game Legends of Arborea

dpac-dialoguegan

This repo implements GAN-based models for Dialogue Generation (DP-GAN, SeqGAN, and our own proposed DPAC-GAN)

encoder-decoder

Encoder-decoder model with attention (Luong), with two LSTM layers with 500 hidden units on both encoder and decoder side. The vocabulary size on both source (english) and target side (Dutch) is 50000. The model is trained on the train part of the TED dataset (https://wit3.fbk.eu/mt.php?release=2017-01-trnmted), maximum sequence length 50.

jsbaan.github.io

kennissystemen

Prolog-based Knowledge Systems on toy problems

machine

Codebase for the i-machine-think project, a modular and fully tested Pytorch implementation for seq2seq models.

natural-gradient-rl

Project for the course Reinforcement Learning about natural gradients

neurodriver

Project on autonomous racing in the Torcs simulator using RNNs, NeuroEvolution of Augmenting Topologies (NEAT) and Swarm Intelligence for the course Computational Intelligence

nlg-uncertainty-probes

Code to probe Natural Language Generation (NLG) systems for their uncertainty, and to evaluate it against human production variability (i.e., aleatoric uncertainty)

nlp1

Implements oldschool NLP methods for the course Natural Language Processing 1

nlp2

Implements IBM1 & IBM 2 with Variational Bayes for the course Natural Language Processing 2

transformer-from-scratch

Well documented, unit tested, type checked and formatted implementation of a vanilla transformer - for educational purposes.