Is any multi discrete action example for PPO or other algorithms?

Hi,
I cannot find multi discrete action example. May someone point out this for me? Thanks in advance.

@James_Liu

How about this one?

1 Like

@mannyv Thanks. I will try it.

@James_Liu,

The more general answer is if you have an environment that defines a multidiscrete space there is not really anything special you have to do. Rllib will support it automatically. This assumes the algorithm you choose is also compatible with it. For example, PPO is but DQN is not.

Welcome to the forum by the way. I forgot to include that in my first response.

1 Like

@mannyv
When running test_lstms.py, I got the unknown config parameter below. Should I configure other things?

Exception: Unknown config parameter _disable_action_flattening
(pid=24784) 2022-01-18 12:32:40,954 ERROR worker.py:428 – Exception raised in creation task: The actor died because of an error raised in its creation task, ray::PPO.init() (pid=24784, ip=192.168.1.5)
(pid=24784) File “/home/james/anaconda3/envs/py38/lib/python3.8/site-packages/ray/rllib/agents/trainer_template.py”, line 136, in init
(pid=24784) Trainer.init(self, config, env, logger_creator)
(pid=24784) File “/home/james/anaconda3/envs/py38/lib/python3.8/site-packages/ray/rllib/agents/trainer.py”, line 592, in init
(pid=24784) super().init(config, logger_creator)
(pid=24784) File “/home/james/anaconda3/envs/py38/lib/python3.8/site-packages/ray/tune/trainable.py”, line 105, in init
(pid=24784) self.setup(copy.deepcopy(self.config))
(pid=24784) File “/home/james/anaconda3/envs/py38/lib/python3.8/site-packages/ray/rllib/agents/trainer_template.py”, line 146, in setup
(pid=24784) super().setup(config)
(pid=24784) File “/home/james/anaconda3/envs/py38/lib/python3.8/site-packages/ray/rllib/agents/trainer.py”, line 700, in setup
(pid=24784) self.config = self.merge_trainer_configs(self._default_config, config,
(pid=24784) File “/home/james/anaconda3/envs/py38/lib/python3.8/site-packages/ray/rllib/agents/trainer.py”, line 1344, in merge_trainer_configs
(pid=24784) return deep_update(config1, config2, _allow_unknown_configs,
(pid=24784) File “/home/james/anaconda3/envs/py38/lib/python3.8/site-packages/ray/tune/utils/util.py”, line 264, in deep_update
(pid=24784) raise Exception("Unknown config parameter {} ".format(k))
(pid=24784) Exception: Unknown config parameter _disable_action_flattening