NVIDIA PhysicsNeMo (formerly Modulus)
Open-source deep-learning framework for physics-ML models (neural operators, GNNs, PINNs).
Release timeline (newest first, grouped by year and version)
2026 6 entries
2.x
- v2.2.2 release
PhysicsNeMo General Release v2.2.2 ### Fixed * Aligns documentation with the PyPi name
source: github-release - v2.2.1 release
PhysicsNeMo General Release v2.2.1 ### Fixed * Allows installation on all Python 3.14 patch releases while continuing to exclude Python 3.15 and later. * Keeps tensorclass mesh API annotations introspectable under Python 3.14's deferred annotation evaluation.
source: github-release - v2.2.0 release
PhysicsNeMo General Release v2.2.0 ### Added - Adds `ShardTensor` support for GeoTransolver and FLARE models. - Adds zarr save/load for `Mesh` and `DomainMesh` via tensordict's zarr storage backend: `physicsnemo.mesh.io.to_zarr` / `from_zarr`, with training-appropriate chunking a
source: github-release - v2.1.1 release
PhysicsNeMo General Release v2.1.1 ### Changed * Refactors GeoTransolver, Flare, and some components of transolver to be enabled for 2D and 3D cases.
source: github-release - v2.1.0 release
PhysicsNeMo General Release v2.1.0 ### Added - Adds GLOBE model (`physicsnemo.experimental.models.globe.model.GLOBE`), including new variant that uses a dual tree traversal algorithm to reduce the complexity of the kernel evaluations from O(N^2) to O(N). - Adds GLOBE AirFRANS exa
source: github-release - v2.0.0 releasemajorMajor version 2
PhysicsNeMo General Release v2.0.0 📝 NVIDIA PhysicsNeMo v2.0 contains significant reorganization of all the features, with easier installation and integration to external packages. See the [migration guide](https://github.com/NVIDIA/physicsnemo/blob/main/v2.0-MIGRATION-GUIDE.md)
source: github-release
2025 6 entries
1.x
- v1.3.0 release
PhysicsNeMo General Release v1.3.0 ### Added - Added mixture_of_experts for weather example in physicsnemo.examples.weather. **⚠️Warning:** - It uses experimental DiT model subject to future API changes. Added some modifications to DiT architecture in physicsnemo.experimental.mod
source: github-release - v1.2.0 release
PhysicsNeMo General Release v1.2.0 ### Added - Diffusion Transformer (DiT) model. The DiT model can be accessed in `physicsnemo.experimental.models.dit.DiT`. **⚠️Warning:** - Experimental feature subject to future API changes. - Improved documentation for diffusion models and dif
source: github-release - v1.1.1 release
PhysicsNeMo (Core) General Release v1.1.1 (patch to v1.1.0) ### Fixed - Fixed an inadvertent change to the deterministic sampler 2nd order correction
source: github-release - v1.1.0 release
PhysicsNeMo (Core) General Release v1.1.0 ### Added - Added ReGen score-based data assimilation example - General purpose patching API for patch-based diffusion - New positional embedding selection strategy for CorrDiff SongUNet models - Added Multi-Storage Client to allow checkp
source: github-release - v1.0.1 release
PhysicsNeMo (Core) General Release v1.0.1 ### Added - Added version checks to ensure compatibility with older PyTorch for distributed utilities and ShardTensor ### Fixed - `EntryPoint` error that occured during physicsnemo checkpoint loading
source: github-release - v1.0.0 releasemajorMajor version 1
PhysicsNeMo (Core) General Release v1.0.0 ### Added - DoMINO model architecture, datapipe and training recipe - Added matrix decomposition scheme to improve graph partitioning - DrivAerML dataset support in FIGConvNet example. - Retraining recipe for DoMINO from a pretrained mode
source: github-release
2024 5 entries
0.x
- v0.9.0 release
Modulus (core) general release v0.9.0 ## Added - FIGConvUNet model and example. - The Transolver model. - The XAeroNet model. - Incoporated CorrDiff-GEFS-HRRR model into CorrDiff, with lead-time aware SongUNet and cross entropy loss. ## Changed - Refactored EDMPrecondSRV2 precond
source: github-release - v0.8.0 release
Modulus (core) general release v0.8.0 ## Added - Graph Transformer processor for GraphCast/GenCast. - Utility to generate STL from Signed Distance Field. - Metrics for CAE and CFD domain such as integrals, drag, and turbulence invariances and spectrum. - Added gradient clipping t
source: github-release - v0.7.0 release
Modulus (core) general release v0.7.0 ## Added - Code logging for CorrDiff via Wandb. - Augmentation pipeline for CorrDiff. - Regression output as additional conditioning for CorrDiff. - Learnable positional embedding for CorrDiff. - Support for patch-based CorrDiff training and
source: github-release - v0.6.0 release
Modulus (core) general release v0.6.0 ## Added - Added citation file - Link to the CWA dataset - ClimateDatapipe: an improved datapipe for HDF5/NetCDF4 formatted climate data - Performance optimizations to CorrDiff - Physics-Informed Nonlinear Shallow Water Equations example - Wa
source: github-release - v0.5.0 release
Modulus (core) general release v0.5.0 ## Added - Distributed process group configuration mechanism. - DistributedManager utility to instantiate process groups based on a process group config. - Helper functions to facilitate distributed training with shared parameters. - Brain an
source: github-release
2023 6 entries
0.x
- v0.4.0 release
Modulus (core) general release v0.4.0 ## Added - Added Stokes flow dataset - An experimental version of SFNO to be used in unified training recipe for weather models. - Added distributed FFT utility. - Added ruff as a linting tool. - Ported utilities from Modulus Launch to main p
source: github-release - v0.3.0 release
Modulus (core) general release v0.3.0 ## Added - Added distributed utilities to create process groups and orthogonal process groups - Added distributed AFNO model implementation - Added distributed utilities for communication of buffers of varying size per rank - Added instructio
source: github-release - v0.2.1 release
Modulus (core) hotfix release v0.2.1 ## Fixed - Added a workaround fix for the CUDA graphs error in multi-node runs ## Security - Update certifi package version
source: github-release - v0.2.0 release
Modulus (core) general release v0.2.0 ## Added - Added a CHANGELOG.md - Added build support for internal DGL - 4D Fourier Neural Operator model - Ahmed body dataset ## Changed - DGL install changed from pypi to source - Updated SFNO to add support for super resolution, flexible c
source: github-release - v0.1.0 releasemajorFirst release
Modulus (core) general release v0.1.0 (Initial OSS release)
source: github-release