git clone https://github.com/hp-transfer/ht_optimizers.git
git clone https://github.com/hp-transfer/ht_benchmarks.git
git clone https://github.com/hp-transfer/htaa_experiments.git
cd htaa_experiments
# Activate virtual / conda environment [-c conda-forge swig gcc_linux-64 gxx_linux-64]
poetry install
<EXPERIMENT_GROUP>/ << Logical group of experiments
├── cluster_oe_<BENCHMARK>/ << Output and error related to benchmark
├── <BENCHMARK>/ << Each benchmark gets own folder
│ └── <RUNTYPE> << Each type of run gets own folder
│ └── <APPROACH>/ << Each approach gets own folder
│ └── trajectory_id_<ID>/ << Each trajectory gets own folder
│ └── adjustment_id_<ID>/ << Each adjustment type gets own folder
│ └── repeat_<REPETITION>/ << Each repetition gets own folder
│ ├── .hydra/ << Exact configuration files used in this run
│ ├── eval/ << Generalization results per step
│ │ ├── batch_result_<STEP>.pkl
│ ├── train/ << Online results
│ │ ├── batch_result_<STEP>.pkl
│ │ └── trajectory.pkl
│ └── <LOGGER>.log
└── <BENCHMARK>.args << Argfile for benchmark
