datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
PDEBench-1D
Dataset Card for "PDEBench-1D"
More Information needed
PDEBench_2D_diff-reactlegal:
owner: Takamoto, M et al. (https://darus.uni-stuttgart.de/dataset.xhtml?persistentId=doi:10.18419/darus-2986)
license: cc-by-4.0
data_production:
physics: 2D Diffusion-Reaction
type: simulation
script: Converted to PLAID format for standardized usage; no changes to data content.
num_samples:
train: 1000
storage_backend: hf_datasets
plaid:
version: 0.1.12
This dataset was generated with plaid, we refer to this documentation for additional details on how to extract data… See the full description on the dataset page: https://huggingface.co/datasets/Nionio/PDEBench_2D_diff-react.PDEBench-1D-full
Dataset Card for "PDEBench-1D-full"
More Information needed
staatsblad-synth-nl
Synthetic Dutch from the Belgisch Staatsblad
Diverse, fluent Dutch pretraining text synthesized from
guust-franssens/belgisch-staatsblad
(CC0, Belgian official-gazette filings). Adds Belgium/Flanders coverage to Dutch LM pretraining
mixes, where clean Belgian-Dutch prose is otherwise scarce.
The source text is noisy OCR from scanned PDFs, but its metadata (company, juridical form,
act type, city, date) is clean. A local LLM (google/gemma-2-9b-it)
"launders" the OCR + metadata… See the full description on the dataset page: https://huggingface.co/datasets/pdelobelle/staatsblad-synth-nl.PDEBench-1Dfineweb-dutch-edu-mt
FineWeb-Edu Dutch Machine Translated
Machine-translated Dutch text dataset derived from the FineWeb-Edu corpus.
Dataset Details
Source: HuggingFaceFW/fineweb-edu (sample-10BT subset)
Translation: English → Dutch using Unbabel/Tower-Plus-9B
Size: Up to 1.5M samples
Format: Translated text with original metadata
Schema
text: Machine-translated Dutch text
id: Original sample identifier from FineWeb-Edu
url: Source URL
Quality Notice
⚠️ This… See the full description on the dataset page: https://huggingface.co/datasets/pdelobelle/fineweb-dutch-edu-mt.PDEBench_2D_DarcyFlowExample of usage:
import torch
from plaid.bridges import huggingface_bridge as hfb
from torch.utils.data import DataLoader
def reshape_all(batch: dict[str, torch.Tensor]) -> dict[str, torch.Tensor]:
"""Helper function that reshapes the flattened fields into images of sizes (128, 128)."""
batch["diffusion_coefficient"] = batch["diffusion_coefficient"].reshape(
-1, 128, 128
)
batch["flow"] = batch["flow"].reshape(-1, 128, 128)
return batch
# Load the dataset… See the full description on the dataset page: https://huggingface.co/datasets/Nionio/PDEBench_2D_DarcyFlow.PDEBench_2D_SWElegal:
owner: Takamoto, M et al. (https://darus.uni-stuttgart.de/dataset.xhtml?persistentId=doi:10.18419/darus-2986)
license: cc-by-4.0
data_production:
physics: Shallow Water Equations
type: simulation
script: Converted to PLAID format for standardized usage; no changes to data content.
num_samples:
train: 1000
storage_backend: hf_datasets
plaid:
version: 0.1.12
This dataset was generated with plaid, we refer to this documentation for additional details on how to extract data… See the full description on the dataset page: https://huggingface.co/datasets/Nionio/PDEBench_2D_SWE.synth-nl
SYNTH-NL
Dutch language subset of the SYNTH dataset by Pleias and the AI Alliance.
This dataset contains only the Dutch (nl) language samples from the original SYNTH corpus, which comprises synthetic training data generated from Wikipedia and Wikibooks articles.
Source Dataset: PleIAs/SYNTH
License: CDLA-Permissive-2.0
Language: Dutch (nl)
For complete dataset documentation, methodology, and usage guidelines, refer to the original SYNTH repository.
PDEBench_2D_DarcyFlow_beta10.0legal:
owner: Takamoto, M et al. (https://darus.uni-stuttgart.de/dataset.xhtml?persistentId=doi:10.18419/darus-2986)
license: cc-by-4.0
data_production:
physics: 2D Darcy Flow
type: simulation
script: Converted to PLAID format for standardized usage; no changes to data content.
num_samples:
train: 10000
storage_backend: hf_datasets
plaid:
version: 0.1.12
This dataset was generated with plaid, we refer to this documentation for additional details on how to extract data from… See the full description on the dataset page: https://huggingface.co/datasets/Nionio/PDEBench_2D_DarcyFlow_beta10.0.nemotron-dutch-mt
Nemotron Post-Training Dataset (Dutch Translation)
Machine-translated Dutch version of NVIDIA's Nemotron Post-Training Dataset, specifically the chat conversations.
Dataset Details
Source: nvidia/Nemotron-Post-Training-Dataset-v2 (chat split)
Translation: English → Dutch using Unbabel/Tower-Plus-9B
Size: 445,287 conversations with 1,327,548 total messages
Format: Conversational data with original structure preserved
Dataset Statistics
Total… See the full description on the dataset page: https://huggingface.co/datasets/pdelobelle/nemotron-dutch-mt.PDEBench_2D_DarcyFlow_beta1.0legal:
owner: Takamoto, M et al. (https://darus.uni-stuttgart.de/dataset.xhtml?persistentId=doi:10.18419/darus-2986)
license: cc-by-4.0
data_production:
physics: 2D Darcy Flow
type: simulation
script: Converted to PLAID format for standardized usage; no changes to data content.
num_samples:
train: 10000
storage_backend: hf_datasets
plaid:
version: 0.1.12
This dataset was generated with plaid, we refer to this documentation for additional details on how to extract data from… See the full description on the dataset page: https://huggingface.co/datasets/Nionio/PDEBench_2D_DarcyFlow_beta1.0.pde-llm-eval-cross-representation-dataset
pde-llm-eval-cross-representation-dataset
Cross-Representation Dataset. 32 physical systems, one row each, with FOUR representations of every system -- source code, natural-language description, governing equation, and numerical trajectory -- each given in its correct form and in its corrupted form. Code additionally comes with real and obfuscated identifiers, and the trajectory with four kinds of corruption (randomly generated values, values permuted within the original… See the full description on the dataset page: https://huggingface.co/datasets/bermaneh/pde-llm-eval-cross-representation-dataset.pde-llm-eval-code-perturbation-dataset
pde-llm-eval-code-perturbation-dataset
Code Perturbation Dataset (final). 256 PDE solver implementations: 64 base programs (32 physically valid, 32 with an injected bug that still runs to completion) expanded by 4 lexical perturbation conditions each -- original comments, comments removed, comments swapped in from a different implementation, and descriptive identifiers replaced by meaningless placeholders. The perturbations change the lexical surface only; executable behaviour… See the full description on the dataset page: https://huggingface.co/datasets/bermaneh/pde-llm-eval-code-perturbation-dataset.PDEBench_2D_DarcyFlow_beta0.01legal:
owner: Takamoto, M et al. (https://darus.uni-stuttgart.de/dataset.xhtml?persistentId=doi:10.18419/darus-2986)
license: cc-by-4.0
data_production:
physics: 2D Darcy Flow
type: simulation
script: Converted to PLAID format for standardized usage; no changes to data content.
num_samples:
train: 10000
storage_backend: hf_datasets
plaid:
version: 0.1.12
This dataset was generated with plaid, we refer to this documentation for additional details on how to extract data from… See the full description on the dataset page: https://huggingface.co/datasets/Nionio/PDEBench_2D_DarcyFlow_beta0.01.fineweb-german-edu-mtPDEBench_2D_DarcyFlow_beta0.1legal:
owner: Takamoto, M et al. (https://darus.uni-stuttgart.de/dataset.xhtml?persistentId=doi:10.18419/darus-2986)
license: cc-by-4.0
data_production:
physics: 2D Darcy Flow
type: simulation
script: Converted to PLAID format for standardized usage; no changes to data content.
num_samples:
train: 10000
storage_backend: hf_datasets
plaid:
version: 0.1.12
This dataset was generated with plaid, we refer to this documentation for additional details on how to extract data from… See the full description on the dataset page: https://huggingface.co/datasets/Nionio/PDEBench_2D_DarcyFlow_beta0.1.PDEBench_2D_DarcyFlow_beta100.0legal:
owner: Takamoto, M et al. (https://darus.uni-stuttgart.de/dataset.xhtml?persistentId=doi:10.18419/darus-2986)
license: cc-by-4.0
data_production:
physics: 2D Darcy Flow
type: simulation
script: Converted to PLAID format for standardized usage; no changes to data content.
num_samples:
train: 10000
storage_backend: hf_datasets
plaid:
version: 0.1.12
This dataset was generated with plaid, we refer to this documentation for additional details on how to extract data from… See the full description on the dataset page: https://huggingface.co/datasets/Nionio/PDEBench_2D_DarcyFlow_beta100.0.sigs-symbolic-pde-corpus
SIGS Grammar Production Corpus
This dataset contains the one-hot grammar-production sequences used to train the
Grammar-VAE in SIGS: Neuro-Symbolic AI for Analytical Solutions of
Differential Equations (Oikonomou et al., ICML 2026).
Structure
Each row contains:
inputs: a float32 tensor shaped [grammar productions, sequence length];
labels: the corresponding integer production indices shaped
[sequence length].
The deterministic default split uses seed 42 with 70%… See the full description on the dataset page: https://huggingface.co/datasets/oroikono/sigs-symbolic-pde-corpus.pde_real_sftThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v3.0",
"robot_type": "fr3_agilex",
"total_episodes": 443,
"total_frames": 237678,
"total_tasks": 11,
"chunks_size": 1000,
"fps": 50,
"splits": {
"train": "0:443"
},
"data_path": "data/chunk-{chunk_index:03d}/file-{file_index:03d}.parquet",
"video_path": "videos/{video_key}/chunk-{chunk_index:03d}/file-{file_index:03d}.mp4"… See the full description on the dataset page: https://huggingface.co/datasets/jmarangola/pde_real_sft.fineweb-dutch-synthetic-mt
FineWeb Dutch Synthetic MT
Machine-translated Dutch text dataset derived from the Aleph-Alpha GermanWeb synthetic corpus.
Dataset Details
Source: Aleph-Alpha/Aleph-Alpha-GermanWeb (synthetic split)
Translation: German → Dutch using Unbabel/Tower-Plus-9B
Size: ~300k+ samples (subset of 398M total)
Format: Plain text with sample IDs
Schema
text: Machine-translated Dutch text
id: Original sample identifier
Quality Notice
⚠️ This is… See the full description on the dataset page: https://huggingface.co/datasets/pdelobelle/fineweb-dutch-synthetic-mt.pde-probe-pilot-probe-pooled-mean-pool-v1
pde-probe-pilot-probe-pooled-mean-pool-v1
LOGO-CV pooled linear probe results (mean_pool), all labels × all layers. 128 rows, 8 mod_types.
Dataset Info
Rows: 270
Columns: 17
Columns
Column
Type
Description
label
Value('large_string')
Target label probed (pde_class, process_*, method_*, phys_valid)
layer
Value('large_string')
Transformer layer index (0=embedding, 1-28=transformer) or 'bow'
pool
Value('large_string')
Pooling strategy (mean_pool)… See the full description on the dataset page: https://huggingface.co/datasets/rosubramanian/pde-probe-pilot-probe-pooled-mean-pool-v1.wikipedia_summaries
Contents
This dataset contains text summaries of 150 topics extracted from Wikipedia. The topics range from simplistic (e.g. sun) to somewhat specific (e.g. machine learning).
Usage
This dataset was created for use with the book Real-Time Machine Learning by Prema Roman and Patrick Deziel and its official github repository
math-pdefakeverifier-datasetpde-probe-pilot-probe-pooled-last-tok-v1
pde-probe-pilot-probe-pooled-last-tok-v1
LOGO-CV pooled linear probe results (last_tok), all labels × all 29 layers. 128 rows, 8 mod_types. Complete.
Dataset Info
Rows: 270
Columns: 17
Columns
Column
Type
Description
label
Value('large_string')
Target label probed (pde_class, process_*, method_*, phys_valid)
layer
Value('large_string')
Transformer layer index (0=embedding, 1-28=transformer) or 'bow'
pool
Value('large_string')
Pooling strategy… See the full description on the dataset page: https://huggingface.co/datasets/rosubramanian/pde-probe-pilot-probe-pooled-last-tok-v1.mediflow-mtfakeverifier-llama-datasetRegular_and_Singular_Stochastic_PDEs