actively being developed
.
├── config
│ ├── cleanrl
│ └── experiment
└── improve
├── cleanrl
├── purejaxrl
│ └── experimental
│ └── s5
├── simpler_mod
└── wrappers
- install SIMPLER
- install Octo
piip install -e .
python main.py job/wandb=dont algo=sac train.use_train=True
- a/b selects config group b from parent group a
- a.b sets attribute b of group a
python main.py -m +exp=may29_sac
-mflag is needed for multirun experiments
- SB3
python improve/sb3/main.py
- CleanRL
python improve/cleanrl/ppo.py
TBD
TBD
TBD
TBD