datasets
Training and evaluation data, with the modality, task and licence stated up front. Listed live from the Hugging Face Hub.
forge-industrial-control-scenarios
Forge Industrial Control and Telemetry Traces
Deterministic synthetic traces spanning device ingress, signature/quality/range
failures, offline store-and-forward, local inference, sequential agent review,
L0–L4 policy outcomes, electrolyser ramp sequences and multivariate telemetry
anomalies.
No operational plant data, customer data, secrets or real equipment identifiers are included. machine.press-03 and every measurement are fictitious.
Files… See the full description on the dataset page: https://huggingface.co/datasets/sankalpsthakur/forge-industrial-control-scenarios.forge-pump-digital-twin-synthetic
Forge Pump Digital Twin — Synthetic
A deterministic, clean-room tabular baseline for pump surrogate modelling, edge-runtime conformance, and advisory anomaly examples. It contains 40,000 synthetic rows split into 28,000 train, 6,000 validation, and 6,000 test rows.
This dataset contains no plant telemetry, customer data, equipment identifiers, CAD, BOMs, nameplates, vendor curves, or values copied from a private repository. Every constant is an illustrative engineering proxy. It… See the full description on the dataset page: https://huggingface.co/datasets/sankalpsthakur/forge-pump-digital-twin-synthetic.sequential-forgetting-benchmark
Sequential Forgetting Benchmark
What does sequential fine-tuning do to what a model already learned? This
dataset is a results ledger with receipts: every row of
results/results.csv links to the raw run file it came from
(results/raw/), every transcription is hand-checked (results/PROVENANCE.md),
and invalid runs are disclosed, not deleted.
Seeded from ModelBrew's archival continual-learning runs (2026). Community
submissions welcome — see protocol/PROTOCOL.md.… See the full description on the dataset page: https://huggingface.co/datasets/ModelBrew/sequential-forgetting-benchmark.