History

0.11

refactored Scaler, Potential, and Trainer
optimized Kalman Filter JIT kernels
separated scaler and model parameters from potential
bug fixes

0.10

added an optimized implementation (100x faster) of the ACSF gradient calculation

0.9

required adjustments on the base codes and example notebooks to work with the recent version of JAX
added autodiff gradient method to the LJ force calculation kernel

0.8

Changed package name to Pantea

0.7

Refactored structure including design, documentation, and performance (JIT kernels)
Refactored RuNNer dataset

0.6

Implemented a Molecular dynamics (MD) simulator
Monte-Carlo (MC) simulator

0.5

Implemented Kalman filter trainer

0.4

Applied extensive refactoring
Replaced PyTorch main dependency with JAX
First release on PyPI.

0.3

JAX optimizations applied to the ACSF descriptor

0.2

Some optimizations using torch.jit.script

0.1

Primary implementation and validation