IBM / simulai

A toolkit with data-driven pipelines for physics-informed machine learning.

https://ibm.github.io/simulai/

IBM/simulai Issues

Add Docstrings to Simulation, Normalization, Metrics, and Mesh Modules
Closed a year ago1
Doctrings for Optimization, SymbolicOperator and Losses
Closed 7 months ago1
Roadmap for SimulAI
Updated 6 months ago
Wavelet activation is not running on CUDA devices
Closed 6 months ago
NIF is not converging.
Updated 7 months ago
SPFile must also save class defined template methods.
Updated 10 months ago
MultiScaleAutoencoder also should support vanilla autoencoders
Closed a year ago
Python 3.11 support
Closed a year ago4
Python version limitation
Closed a year ago2
Verbose printing is not properly operating
Updated a year ago
Batchwise optimization takes longer times than expected to run
Closed a year ago3
Causality preserving must be generic and externally defined.
Closed a year ago
Is the SciPyInterface really communicating gradients to the torch.nn.Module objects ?
Updated a year ago
Inputs keys for Physics-Informed Neural Operators could contain variables from any branch
Closed a year ago
Batch-normalization must be considered on CNN architectures.
Closed a year ago
Integration with SciPy optimizers
Closed a year ago
Trainable parameters are not being updated
Closed a year ago
NumPy docstrings for the directory simulai/models
Closed a year ago2
Replicating the structure of PIRMSE to OPIRMSE
Updated a year ago
Creating units tests
Updated a year ago
SymbolicOperator should support to receive def functions.
Closed a year ago1
Model Checkpointing is missing.
Closed a year ago1
Filtering losses in PINNs
Closed a year ago
Some variables, as energy and concentration need to preserve positivity otherwise they became physically inconsistent. When these variables are represented via Autoencoders and PCA the way for enforcing positivity is the hard positivity limiting. It already works for PCA, but it still needs to be tested with NN Autoencoders.
Closed a year ago
Updating to torch >=1.13
Closed a year ago
📌 ISSUES TEMPLATE
Updated a year ago
Sanity tests for Hard Positivity Limiting
Closed a year ago
Sanity tests for normalization.
Closed a year ago
Unknown license found
Closed a year ago3
Help: Missing Class in PyTorch Residuals Module
Closed a year ago6
Optimizer needs to be completely compatible with multi-GPU jobs
Closed a year ago
Cleaning unused or merged branches.
Closed a year ago1
numpy.float deprecated
Closed a year ago3
🦋 Suggest: using black and isort before committing code
Closed a year ago1
The dense bottlenecks of automatic generated AEs could be optionally shallow
Closed a year ago1
Add NumPy style Docstrings to SimulAI packages
Closed a year ago7
Extend automatic generation for autoencoders Variational and Koopman
Closed a year ago
📌 Enhancement Request: Develop Project Roadmap
Closed a year ago1
📌 Problem: Lack of Detailed Release Information 🛠
Closed a year ago1
Sanity tests for MoE Pool
Closed a year ago
📌 Problem: Excessive Commits and Poor Release Output 🛠
Closed a year ago1
Error: Unknown target name: "mask".sphinx
Closed a year ago1
Missing documentation for module parallel
Closed a year ago2
UserWarning: Hamilatonian sampling is in EXPERIMENTAL stage.
Closed a year ago
Error when trying to perform internal boundary product with numpy array
Closed a year ago3
Sphinx Build Successful but with 146 Warnings
Closed a year ago7
Docs still ignore third level of directories ("sub-sub-packages").
Closed 2 years ago12
Dedicated requirements.txt for the examples
Closed 2 years ago1
Rename simulai/regression/pytorch to simulai/regression/_pytorch
Closed 2 years ago1
Docs compilation is failling because of dependency sphinx_theme is not found
Closed 2 years ago2