datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
Peregrine-Dataset-v2023-11
Peregrine v2023-11 — Layer-wise L-PBF Imaging Dataset
A HuggingFace-formatted conversion of the Peregrine v2023-11 dataset released by Oak Ridge National Laboratory's (ORNL) Manufacturing Demonstration Facility (MDF). The dataset contains layer-wise in-situ imaging, anomaly segmentation masks, laser scan paths, and process sensor data from 5 Laser Powder Bed Fusion (L-PBF) builds of stainless steel 316L.
Original dataset: Layer-wise Imaging Dataset from Powder Bed Additive… See the full description on the dataset page: https://huggingface.co/datasets/ppak10/Peregrine-Dataset-v2023-11.NIST-In-Situ-IN625-LPBF-OverhangsAdditive-Manufacturing-Benchmark
Additive Manufacturing Benchmark
A benchmark dataset for evaluating knowledge of additive manufacturing (AM) processes, derived from graduate-level coursework at Carnegie Mellon University.
Configurations
general_knowledge_multiple_choice
Multiple-choice questions covering various AM processes with explanations.
Column
Description
source
Source homework assignment (e.g. cmu_24_633_2023/homework_1_exone)
process
AM process type (e.g. Binder Jet… See the full description on the dataset page: https://huggingface.co/datasets/ppak10/Additive-Manufacturing-Benchmark.AdditiveLLM2-OA
AdditiveLLM2-OA Dataset
Open Access journal articles (up to February 2026) used in domain adapting
pretraining and instruction tuning for AdditiveLLM2.
Dataset Split by Journal
text
images
vit
Vocabulary Overlap
Pairwise Jaccard similarity of word-level vocabularies (lowercase, 3+ letter tokens) across the four source journals. Run info/vocabulary/vocabulary_overlap.py to reproduce.
Top Phrases by Journal
Most frequent bigrams and… See the full description on the dataset page: https://huggingface.co/datasets/ppak10/AdditiveLLM2-OA.FLOW-3D-Single-Track
FLOW-3D-Single-Track
This dataset consists of LPBF (Laser Powder Bed Fusion) melt pool simulations performed using FLOW-3D. Each row is one timestep from one simulation case, containing process parameters, mesh metadata, a per-timestep field volume, and image previews.
Dataset Configs
Load a specific field config:
from datasets import load_dataset
ds = load_dataset("ppak10/FLOW-3D-Single-Track", "temperature")
Available configs:
Config
Volume Shape… See the full description on the dataset page: https://huggingface.co/datasets/ppak10/FLOW-3D-Single-Track.Agentic-SLS-Telemetry
Inova-Mk1-Telemetry
Time-aligned printer-state recordings from Inova Mk1 SLS 3D print runs. One row per 10 Hz tick — the recorder's /state/snapshot poll — with the full sensor state snapshot (~64 columns: temperatures, position, power, lights) on every row, the nearest camera frame embedded inline when one fell within the prior 100 ms window, and any 1 kHz position-stream samples from that window collected as a nested list.
25 parquet files across builds spanning 2026-05 through… See the full description on the dataset page: https://huggingface.co/datasets/ppak10/Agentic-SLS-Telemetry.Gelsight-Welds
Gelsight-Welds
Paired tactile (GelSight Mini) and top-down RGB (Intel RealSense D555) captures of weld beads on a flat sample plate. Each sample (session) is scanned with a 10×10 grid of GelSight touches; the D555 takes one overhead shot per session, which we crop into per-tile RGB images aligned with each touch.
Loading
from datasets import load_dataset
ds = load_dataset("Gelsight-Welds", split="all") # uses config "v2" by default; no train/val/test split is… See the full description on the dataset page: https://huggingface.co/datasets/ppak10/Gelsight-Welds.SS316L_small_10_micronmelt-pool-classificationnr_ppar_gamma_tox21-multimodalrecord-blueThis dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "so100_follower",
"total_episodes": 3,
"total_frames": 2626,
"total_tasks": 1,
"total_videos": 6,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 30,
"splits": {
"train": "0:3"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/ppankti81/record-blue.loki-eval-datasetpp_allkomi-yazva-russian-parallel
komi-yazva-russian-parallel
This repository contains a sentence-level parallel corpus for Komi-Yazva to Russian translation.
Dataset structure
Split: train
Rows: 457
Stories: 74
Columns:
row_id: stable row identifier from the local pipeline
story_id: integer story identifier from the source spreadsheet
paragraph_id: integer paragraph identifier from the source spreadsheet
source_row_number: original 1-based spreadsheet row number (including header offset)
komi_yazva:… See the full description on the dataset page: https://huggingface.co/datasets/pparshakov/komi-yazva-russian-parallel.record-test3This dataset was created using LeRobot.
Dataset Structure
meta/info.json:
{
"codebase_version": "v2.1",
"robot_type": "so101_follower",
"total_episodes": 1,
"total_frames": 312,
"total_tasks": 1,
"total_videos": 1,
"total_chunks": 1,
"chunks_size": 1000,
"fps": 30,
"splits": {
"train": "0:1"
},
"data_path": "data/chunk-{episode_chunk:03d}/episode_{episode_index:06d}.parquet",
"video_path":… See the full description on the dataset page: https://huggingface.co/datasets/ppama/record-test3.meeting-to-json-kopp-attach-audioNIST-AMMT-Overhang-Part-X4
Config Notes:
DAQ-XYPT-MPM
Data obtained from DAQ and XYPT are longer than that of MPM as such rows without
trigger activated will have the following changes to save space:
xypt_trigger > 0:
Shapes:
mpm_frame -> (1, 1)
mpm_approx_temp -> (1, 1)
xypt_trigger == 0:
Shapes:
mpm_frame -> (120, 120)
mpm_approx_temp -> (120, 120)
pp-attach-audio-taggedMelt-Pool-Thermal-Imagesnr_ppar_gamma_tox21
Dataset Details
Dataset Description
Tox21 is a data challenge which contains qualitative toxicity measurements
for 7,831 compounds on 12 different targets, such as nuclear receptors and stress
response pathways.
Curated by:
License: CC BY 4.0
Dataset Sources
corresponding publication
data source
assay name
Citation
BibTeX:
@article{Huang2017,
doi = {10.3389/fenvs.2017.00003},
url = {https://doi.org/10.3389/fenvs.2017.00003},
year = {2017}… See the full description on the dataset page: https://huggingface.co/datasets/jablonkagroup/nr_ppar_gamma_tox21.residual_heat_50_micronsRAISE-LPBF-Part-X16residual_heatresidual_heat_large