CODES
Overview
Benchmark
Documentation
Config Maker
Configuration
This page serves as an interacitve config generator.
Config Generator
Name of Configuration
Surrogates to include
Fully connected Neural Network
Deep Operator Network
Latent Neural ODE
Latent Polynomial Network
Add model
Compare models with each other
Use optimal model hyperparameters
Dataset
Take the logarithm of the data
Standardize
Min-Max Normalize [-1, 1]
None
Misc
Verbose output
Models to train
Interpolation
Interpolation
Use every n-th timestep
Extrapolation
Extrapolation
Cutoff at timestep n
Performance on trainingdata subset
Sparse training data
Take every n-th sample
Uncertainty Quantification
Uncertainty Quantification
Ensemble size
Batch size scaling
Batch size scaling
Batch sizes to evaluate
Evaluations during benchmark
Save and plot losses
Save and plot dynamic accuracy
Save and plot the timing
Measure Compute performance
CODES
CODES