New manual for MLatom@XACS
We are happy to announce that our brand-new manual for MLatom@XACS is now online at http://mlatom.com/manual/.
Our new manual is re-designed to make it much easier to find what you need. We also greatly updated it with additional information, useful tips, and more examples that allow the user to quickly start using MLatom for the required calculations.
The manual starts with a bird-eye overview of all MLatom capabilities, with focus on simulation types, machine learning tasks, and data set operations:
Simulations
- single-point calculations
- geometry optimizations (minima and transition states, IRC)
- frequencies & thermochemistry
- UV/vis spectra (ML-NEA)
- simulations with pre-trained models (AIQM1, ANI-1ccx, etc.)
- simulations with user-trained models
- two-photon absorption cross sections (ML-TPA)
- Coming soon:
- molecular dynamics
- IR spectra
Learning
- training popular ML models (KREG, ANI, sGDML, PhysNet, DPMD, GAP-SOAP, KRR-CM)
- training generic ML models (kernel ridge regression with many kernels)
- optimizing hyperparameters
- evaluating ML models (also with learning curves)
- Δ-learning
- self-correction
Data
- converting XYZ coordinates to molecular descriptor (RE, Coulomb matrix, …)
- analyzing data sets
- sampling (random, structure-based, farthest-point) and splitting datasets
As you see, MLatom support a wide range of common tasks for efficient computational chemistry research ranging from accurate geometry optimizations to thermochemical calculations to spectra simulations with many pre-trained machine learning models and also allow the user to easily create and employ their own models. The manual guides the user through all the typical steps and also provides links to many detailed tutorials which we also use for teaching and workshops. Most of the MLatom features can be used on MLatom@XACS cloud withou the need to install MLatom locally. XACS cloud is a free platform going beyond machine learning and offering unique features on its own such as valence bond calculations and energy decomposition analysis.
We also provide timely live support on our Slack workspace and highly recommend all users to join it for more efficient and informal communication with MLatom team and for receiving timely updates.
MLatom is constantly evolving, thanks to contributions of many people and also thanks to your feedback. MLatom is free and open source and benefits from many great open source third-party packages.
Welcome to check out our new MLatom and please do not hesitate to provide feedback. If you think that some information would be useful to include, let us know.
Leave a Reply