Name		Name	Last commit message	Last commit date
parent directory ..
a2c		a2c
a2c_trainer		a2c_trainer
bandits		bandits
cql		cql
cql_trainer		cql_trainer
crossq		crossq
ddpg		ddpg
ddpg_trainer		ddpg_trainer
decision_transformer		decision_transformer
diffusion_bc		diffusion_bc
discrete_sac		discrete_sac
dqn		dqn
dqn_trainer		dqn_trainer
dreamer		dreamer
dreamer_v3		dreamer_v3
expert-iteration		expert-iteration
gail		gail
grpo		grpo
impala		impala
iql		iql
iql_trainer		iql_trainer
media		media
multiagent		multiagent
multiagent_trainer		multiagent_trainer
offline_to_online		offline_to_online
pilco		pilco
ppo-async		ppo-async
ppo		ppo
ppo_trainer		ppo_trainer
redq		redq
reinforce_trainer		reinforce_trainer
rnd		rnd
sac		sac
sac_trainer		sac_trainer
td3		td3
td3_bc		td3_bc
td3_trainer		td3_trainer
tqc		tqc
vla_grpo		vla_grpo
README.md		README.md

README.md

Examples

We provide examples to train the following algorithms:

To run these examples, make sure you have installed hydra:

pip install --upgrade "hydra-core>=1.3,<1.4"

Scripts can be run from the directory of interest using:

python sac.py

or similar. Hyperparameters can be easily changed by providing the arguments to hydra:

python sac.py collector.frames_per_batch=63