software · Global

PhysicsNeMo

Open-source deep-learning framework for building, training, and fine-tuning deep learning models using state-of-the-art Physics-ML methods

Claim this listing

Represent PhysicsNeMo? Send a claim for review so we can discuss listing updates. Claims do not automatically grant ownership or verification.

Data offered docs-verified: reusable PyTorch model architectures for physics ML (e.g. FNO, MeshGraphNet, Transolver, GeoTransolver, DoMINO) · reusable library components and layers · end-to-end training recipes and example workflows (e.g. flood modeling, reservoir surrogates)
Formats docs-verified: Python / PyTorch library (pip install nvidia-physicsnemo)
Access docs-verified: Free and open-source; Apache-2.0 licence (licence per ENTRY metadata, not stated in DOCS)

Use cases llm-inferred: Train a surrogate for flood inundation prediction from existing hydraulic model runs, adapting the flood modeling recipe · Build a reservoir or groundwater-flow surrogate from subsurface simulation outputs, adapting the reservoir simulation recipe · Emulate CFD-style stormwater or culvert flow responses for rapid design sensitivity testing
Audience llm-inferred: ML engineers and data scientists · hydrologists and flood modellers · CFD and computational engineers · structural and geotechnical engineers

Constraints llm-inferred: A modelling framework, not a data source; no UK or hydrological datasets are provided · Requires PyTorch and a GPU/CUDA environment for the documented installs (cu13 and CUDA 12 options) · Requires machine-learning expertise to train, validate and adapt models to site data · DOCS do not state a release cadence or support commitment

Auto-extracted, unverified fields: tools, install, auth. Check upstream documentation.

Verified via GitHub API (2026-10-15). Tools/install/auth auto-extracted from README.

Metadata observed 14/10/2026 · research sweep: manual-2026-10-15 (Chris sources (NVIDIA Earth-2))