Dataset Viewer
Auto-converted to Parquet Duplicate
mixture_T_id
listlengths
3
3
value
float64
-7.47
-3.54
T
float64
293
318
P
float64
100
101
cmp_ids
listlengths
2
2
cmp_mole_fractions
listlengths
2
2
joint_prediction
float64
-7.46
-3.6
joint_residual
float64
-0.01
0.06
joint_nonlinearity
float64
-0
0.27
joint_nonlinearity_std
float64
0.01
0.04
joint_leverage
float64
0.04
0.26
local_prediction
float64
-7.47
-3.61
local_residual
float64
-0.01
0.07
local_nonlinearity
float64
-0
0.28
local_nonlinearity_std
float64
0
0.08
local_leverage
float64
0.13
0.88
source_row_index
int64
60.5k
96.1k
source_repo
stringclasses
1 value
source_split
stringclasses
1 value
source_revision
stringclasses
1 value
mixture_id
stringclasses
4 values
n_atoms
int32
7.1k
18.3k
n_frames
int32
20
20
atom_elements
listlengths
7.1k
18.3k
atom_molecule_id
listlengths
7.1k
18.3k
molecule_smiles
listlengths
2
2
frames
listlengths
20
20
velocities
listlengths
20
20
box_vectors
listlengths
20
20
time_ps
listlengths
20
20
md_status
stringclasses
1 value
md_config_hash
stringclasses
1 value
[ 28, 1576, 313 ]
-6.059315
313.15
101.325
[ 1576, 28 ]
[ 0.0997, 0.9003 ]
-6.048144
-0.011171
0.099534
0.008751
0.255306
-6.047106
-0.012209
0.099535
0.015398
0.860048
96,147
ai4chems/nist-logv
well_fit
51738e7948c770425625c72098bf7edaf18facda
well_fit_96147
7,100
20
["C","C","N","C","C","N","C","C","O","C","C","O","S","O","O","C","F","F","F","H","H","H","H","H","H"(...TRUNCATED)
[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0(...TRUNCATED)
[ "CCn1cc[n+](CCOC)c1.O=S(=O)([O-])C(F)(F)F", "CC(C)O" ]
[[[1.7184079885482788,2.169581890106201,3.078623056411743],[1.8392802476882935,2.24326229095459,3.02(...TRUNCATED)
[[[-0.25411951541900635,-0.4681397080421448,0.8387572765350342],[0.3773663640022278,-0.6760653853416(...TRUNCATED)
[[[4.216837406158447,0.0,0.0],[0.0,4.216837406158447,0.0],[0.0,0.0,4.216837406158447]],[[4.219250202(...TRUNCATED)
[ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 ]
ok
[ 111, 403, 318 ]
-7.466372
318.15
101
[ 111, 403 ]
[ 0.45, 0.55 ]
-7.462487
-0.003885
-0.00115
0.006439
0.03629
-7.465396
-0.000975
-0.001591
0.000752
0.127161
60,462
ai4chems/nist-logv
well_fit
51738e7948c770425625c72098bf7edaf18facda
well_fit_60462
9,450
20
["C","C","N","C","C","C","C","C","C","H","H","H","H","H","H","H","H","H","H","H","C","C","N","C","C"(...TRUNCATED)
[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0(...TRUNCATED)
[ "CCNc1ccccc1", "COC(C)(C)C" ]
[[[3.1424367427825928,2.215662717819214,2.3011386394500732],[2.9938862323760986,2.1628968715667725,2(...TRUNCATED)
[[[0.02593938075006008,0.2859688699245453,0.643478274345398],[-0.5621137619018555,-0.051631353795528(...TRUNCATED)
[[[4.661276340484619,0.0,0.0],[0.0,4.661276340484619,0.0],[0.0,0.0,4.661276340484619]],[[4.674961090(...TRUNCATED)
[ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 ]
ok
[ 27, 245, 308 ]
-3.544952
308.15
100
[ 245, 27 ]
[ 0.9014, 0.0986 ]
-3.602881
0.057928
0.274429
0.043651
0.259835
-3.614782
0.06983
0.276855
0.080795
0.875823
70,482
ai4chems/nist-logv
well_fit
51738e7948c770425625c72098bf7edaf18facda
well_fit_70482
18,334
20
["C","C","C","C","N","C","C","N","C","C","O","S","O","N","S","O","O","C","F","F","F","C","F","F","F"(...TRUNCATED)
[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0(...TRUNCATED)
[ "CCCCn1cc[n+](C)c1.O=S(=O)([N-]S(=O)(=O)C(F)(F)F)C(F)(F)F", "CO" ]
[[[4.089453220367432,3.7906670570373535,3.147709369659424],[3.9398813247680664,3.7812063694000244,3.(...TRUNCATED)
[[[0.3855403661727905,-0.5647233724594116,0.10711562633514404],[1.0075044631958008,-1.60290610790252(...TRUNCATED)
[[[6.731111526489258,0.0,0.0],[0.0,6.731111526489258,0.0],[0.0,0.0,6.731111526489258]],[[6.703973293(...TRUNCATED)
[ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 ]
ok
[ 34, 1118, 293 ]
-4.848261
293.15
100
[ 1118, 34 ]
[ 0.8001, 0.1999 ]
-4.851713
0.003452
0.187635
0.012068
0.246369
-4.842462
-0.005799
0.19037
0.010559
0.401721
64,881
ai4chems/nist-logv
well_fit
51738e7948c770425625c72098bf7edaf18facda
well_fit_64881
13,200
20
["C","N","C","C","N","N","C","C","C","C","N","C","N","H","H","H","H","H","H","H","H","H","H","H","H"(...TRUNCATED)
[0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0,0(...TRUNCATED)
[ "CN(C)C(=N)[NH+](C)C.c1c[n-]cn1", "CCCO" ]
[[[3.7389650344848633,2.4544458389282227,3.106470823287964],[3.6031653881073,2.5006296634674072,3.09(...TRUNCATED)
[[[0.6837537288665771,-0.24712331593036652,0.594387412071228],[0.47024986147880554,0.401522189378738(...TRUNCATED)
[[[5.009620189666748,0.0,0.0],[0.0,5.009620189666748,0.0],[0.0,0.0,5.009620189666748]],[[4.993514060(...TRUNCATED)
[ 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20 ]
ok

ai4chems/nist-logv-md

Molecular-dynamics frames conditioned on a Hugging Face viscosity dataset derived from CheMixHub's NIST-logV task. Each row carries the source row's labels plus an MD trajectory generated at the same state point ({T, P}). Built by mixtureml-md-data at commit c4bbf50.

Schema

Source columns (passed through from ai4chems/nist-logv:well_fit)

column type description
T float64 temperature (K)
P float64 pressure (kPa)
value float64 log viscosity (experimental target)
cmp_ids list<int64> compound ids in the mixture
cmp_mole_fractions list<float64> mole fractions, aligned with cmp_ids
mixture_T_id list<int64> source (cmp_ids..., round(T)) key
joint_*, local_* float64 cubic-fit diagnostics (see source dataset card)

Provenance columns

column type description
source_repo string source HF repo id
source_split string source split name
source_revision string resolved commit sha of the source repo
source_row_index int64 row index in the source split's parquet

MD columns (added by this pipeline)

column type description
mixture_id string stable id of the form <source_split>_<source_row_index>
n_atoms int32 atoms per frame
n_frames int32 number of frames in frames
atom_elements list<string> element symbol per atom
atom_molecule_id list<int32> species index per atom (indexes molecule_smiles)
molecule_smiles list<string> one SMILES per unique component
frames list<list<list<float32>>> coordinates (n_frames, n_atoms, 3), nanometers
velocities list<list<list<float32>>> atom velocities (n_frames, n_atoms, 3), nm/ps
box_vectors list<list<list<float32>>> box vectors (n_frames, 3, 3), nanometers
time_ps list<float32> per-frame time in picoseconds
md_status string "ok" on success
md_config_hash string hash of force field + protocol + sampler config

Generation

  • Force field: OpenFF Sage 2.2.1 (openff-2.2.1.offxml)
  • Partial charges: NAGL openff-gnn-am1bcc-1.0.0
  • Packing: Packmol, cubic periodic box at 1.0 g/mL initial density
  • Protocol: minimize → 100 ps NVT warmup → 2 ns NPT equilibrate → 5 ns NPT production
  • Timestep: 4 fs equilibrate/production with HMR (H mass = 3.0 amu)
  • Frame sampler: EvenStrideSampler (evenly spaced across production)

Source rows whose SMILES or parameterization cannot be handled by Sage 2.2.1 are omitted. See failed.txt next to the shards for the skip list.

Loading

from datasets import load_dataset

ds = load_dataset("ai4chems/nist-logv-md", split="train")
row = ds[0]
print(row["value"])                          # experimental log viscosity
frames = row["frames"]                       # list of (n_atoms, 3) nm
print(len(frames), "frames,", row["n_atoms"], "atoms")
print(row["source_repo"], row["source_revision"])  # provenance
Downloads last month
30