General reactive machine learning potentials for CHON elements

The data of the article "General reactive machine learning potentials for CHON elements"1. GDB-10-rxn_irc_path.extxyzDescription: Test dataset containing 7,790 structures sampled from 779 elementary reactions in GDB-10 (10 equidistant points per IRC path). Includes energy and force labels computed at the $\omega$B97X-D/def2-TZVP level.2. GDB-17-rxn_irc_path.extxyzDescription: Test dataset containing 1,740 structures sampled from 174 elementary reactions in GDB-17 (10 equidistant points per IRC path). Includes energy and force labels computed at the $\omega$B97X-D/def2-TZVP level.3. GDB-10-rxn_raw.tar.gzDescription: Raw xyz files for the 779 reactions in the GDB-10-rxn set. Contains optimized geometries for Reactants (R), Products (P), and Transition States (TS), along with the full Intrinsic Reaction Coordinate (IRC) trajectories.4. GDB-17-rxn_raw.tar.gzDescription: Raw xyz files for the 174 reactions in the GDB-17-rxn set. Contains optimized geometries for Reactants (R), Products (P), and Transition States (TS), along with the full Intrinsic Reaction Coordinate (IRC) trajectories.