Alex Dunn's Projects
my personal site, live at
atomate is a powerful software for computational materials science and contains pre-built workflows.
An automatic "black-box" yet interpretable engine for predicting materials properties.
Battery evaluation and early prediction
Crystal graph convolutional neural networks for predicting material properties.
[inactive] alex and breanna's cs188 project
[inactive] for homework 1
an ultra-minimal productivity system
Implementation of Dreambooth (https://arxiv.org/abs/2208.12242) with Stable Diffusion (tweaks focused on training faces)
dumb simple (server) monitoring tool
The Fireworks Workflow Management Repo.
An implementation of model parallel autoregressive transformers on GPUs, based on the DeepSpeed library.
Latent representations of materials
Common NLP (text mining) tools for materials science and chemistry, for groups at Lawrence Berkeley National Lab (LBNL) and beyond.
Supplementary Materials for Tshitoyan et al. "Unsupervised word embeddings capture latent knowledge from materials science literature", Nature (2019).
MatBench: Benchmark for materials science property prediction
General training framework for NER models on MatBERT
study structural communities in materials databases
data mining for Materials Science
A repo of examples for the matminer (https://github.com/hackingmaterials/matminer) code
Materials Scholar Website Code
Using applied statistics to predict the outcomes of mixed martial arts bouts.
MODNet: a framework for machine learning materials properties
Platform for materials scientists to contribute and disseminate their materials data through Materials Project
gta iv's niko bellic takes on reddit
Python Materials Genomics (pymatgen) is a robust materials analysis code that defines core object representations for structures and molecules with support for many electronic structure codes. It is currently the core analysis code powering the Materials Project.
plug-n-play black box optimizer for high-throughput computing
random bits of text and notes for quick hacking
A Python Automated Machine Learning tool that optimizes machine learning pipelines using genetic programming.
A website for the UC Berkeley Graduate Data Visualization Contest 2019