metadata
dataset_info:
- config_name: chirality
features:
- name: key
dtype: string
- name: mol_id
dtype: string
- name: en_id
dtype: string
- name: n_conformers
dtype: int64
- name: offset
dtype: int64
- name: mol_blocks
list: string
splits:
- name: train
num_bytes: 333811320
num_examples: 14903
download_size: 96007912
dataset_size: 333811320
- config_name: rotation
features:
- name: key
dtype: string
- name: shard
dtype: int32
- name: n_conformers
dtype: int32
- name: offset
dtype: int64
- name: mol_blocks
list: string
- name: torsion_deg
list: float32
splits:
- name: train
num_bytes: 26639308995
num_examples: 1559779
download_size: 7531603014
dataset_size: 26639308995
- config_name: traj_energies
features:
- name: mol_type
dtype: string
- name: n_frames
dtype: int64
- name: energies
list: float64
splits:
- name: train
num_bytes: 8000212
num_examples: 10
download_size: 1741476
dataset_size: 8000212
- config_name: traj_frames
features:
- name: mol_type
dtype: string
- name: frame_idx
dtype: int32
- name: mol_block
dtype: string
splits:
- name: train
num_bytes: 1393574740
num_examples: 999988
download_size: 345766761
dataset_size: 1393574740
configs:
- config_name: chirality
data_files:
- split: train
path: chirality/train-*
- config_name: rotation
data_files:
- split: train
path: rotation/train-*
- config_name: traj_energies
data_files:
- split: train
path: traj_energies/train-*
- config_name: traj_frames
data_files:
- split: train
path: traj_frames/train-*
3DCS Datasets
This repository hosts multiple 3DCS datasets under separate configs.
Configs
chirality: keys, enantiomers, and MolBlocks with 3D coordinatestraj_frames: per-frame MolBlocks for trajectory datatraj_energies: per-trajectory energy arraysrotation: rotation conformers with MolBlocks and torsion angles
Usage
from datasets import load_dataset
chirality = load_dataset("EscheWang/3dcs", name="chirality")
rotation = load_dataset("EscheWang/3dcs", name="rotation")