STEER Regularization

Installing the steer version of torchdiffeq :

Prerequisites

• Linux or macOS
• Python 3
• CPU or NVIDIA GPU + CUDA CuDNN

Install the bundled version of torchdiffeq

cd torchdiffeq
python setup.py clean
python setup.py build
python setup.py install

STEER: Stiff ODE Experiments

Steer version

cd stiff_ode_experiments
python stiff_ode_demo.py --viz --data_size 120 --version steer --min_length 0.001 --stiffness_ratio 1000 --niters 3000

Standard Neural ODE version

cd stiff_ode_experiments
python stiff_ode_demo.py --viz --data_size 120 --version standard --stiffness_ratio 1000 --niters 3000

Gaussian version

cd stiff_ode_experiments
python stiff_ode_demo.py --viz --data_size 120 --adjoint --version steer --normal_std 0.124 --stiffness_ratio 1000 --niters 3000

STEER: Latent ODE Experiments

MuJoCo Experiment

cd latent_ode
python run_models.py --niters 300 -n 10000 -l 15 --dataset hopper --latent-ode --rec-dims 30 --gru-units 100 --units 300 --gen-layers 3 --rec-layers 3

Physionet

cd latent_ode
python run_models.py --niters 100 -n 8000 -l 20 --dataset physionet --latent-ode --rec-dims 40 --rec-layers 3 --gen-layers 3 --units 50 --gru-units 50 --quantization 0.016 --classif

Human Activity

cd latent_ode
python run_models.py --niters 200 -n 10000 -l 15 --dataset activity --latent-ode --rec-dims 100 --rec-layers 4 --gen-layers 2 --units 500 --gru-units 50 --classif  --linear-classif

FFJORD: CNF experiments

cd ffjord/diagnostics
python approx_error_1d.py --time_length 1
python fig_1_1d_toy.py --save experiments/approx_error_1d --resume experiments/approx_error_1d