AI for Scientific Discovery
The llama-recipes repository is a companion to the Llama 2 model. The repository includes scripts for fine-tuning Llama 2 on text summarisation and question answering, running inference with a fine-tuned Llama 2 model and demo apps to showcase Llama 2 usage with local, cloud, or on-prem deployment.
[ICML'24] Adsorbate Placement via Conditional Denoising Diffusion
Accuracy is not all you need
Torch-native, batchable, atomistic simulations.
PyTorch implementation of EMT potential
🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.
FAIR Chemistry's library of machine learning methods for chemistry
A simulation package of phonon-phonon interaction related properties
An evaluation framework for machine learning models simulating high-throughput materials discovery.
atomate2 is a library of computational materials science workflows
Efficient And Fully Differentiable Extended Tight-Binding
Code for “From Molecules to Materials Pre-training Large Generalizable Models for Atomic Property Prediction”.
Inference code for LLaMA models
A Large Language Model of the CIF format for Crystal Structure Generation
Large language models to generate stable crystals.
Config files for my GitHub profile.
Matbench: Benchmarks for materials science property prediction
An SE(3)-invariant autoencoder for generating the periodic structure of materials [ICLR 2022]
This repository contains implementations and illustrative code to accompany DeepMind publications
A small walkthrough of Facebook's TransCoder Pretrained model for Code Conversion
PyTorch Extension Library of Optimized Scatter Operations
Implementation of Lie Transformer, Equivariant Self-Attention, in Pytorch
Accurate Neural Network Potential on PyTorch
Workflow for creating and analyzing the Open Catalyst Dataset
SchNetPack - Deep Neural Networks for Atomistic Systems