Is this your dataset? Claim it with the Hugging Face account that owns it.

divelab/
PubChemQCR

Name: PubChemQCR
Creator: divelab
License: https://spdx.org/licenses/CC-BY-4.0.html

cc-by-4.0

2506.23008

Views

No views yet

	PubChemQCR








	
	Description

PubChemQCR dataset contains the relaxation trajectory of ~3.5 million small molecules, which can facilitate the development of machine learning interatomic potential (MLIP) models. The relaxation is performed sequentially using PM3, Hartree-Fock, and DFT methods, resulting in a total of 300 million snapshots, 105 million of which are computed using DFT. The dataset is split into two portions, a subset and a full set. Both sets share the same… See the full description on the dataset page: https://huggingface.co/datasets/divelab/PubChemQCR.