Dataset Cards - Alberdice

small-2ag - Download

Environment name	Version	Agents	Action type	Observation size	Reward type
RWARE	Code included in Alberdice repository	2	Discrete	[71]	Dense

Converted from alberdice format to a Vault.

Uid	Episode return mean	Min return	Max return	Transitions	Trajectories	Joint SACo
Expert	7.12 ± 2.07	1.13	12.37	500000	1000	0.99

small-4ag - Download

Environment name	Version	Agents	Action type	Observation size	Reward type
RWARE	Code included in Alberdice repository	4	Discrete	[71]	Dense

Converted from alberdice format to a Vault.

Uid	Episode return mean	Min return	Max return	Transitions	Trajectories	Joint SACo
Expert	9.49 ± 0.84	3.93	12.08	500000	1000	1.00

Environment name	Version	Agents	Action type	Observation size	Reward type
RWARE	Code included in Alberdice repository	6	Discrete	[71]	Dense

Converted from alberdice format to a Vault.

Uid	Episode return mean	Min return	Max return	Transitions	Trajectories	Joint SACo
Expert	10.76 ± 0.68	7.59	12.69	500000	1000	1.00

Environment name	Version	Agents	Action type	Observation size	Reward type
RWARE	Code included in Alberdice repository	2	Discrete	[71]	Dense

Converted from alberdice format to a Vault.

Uid	Episode return mean	Min return	Max return	Transitions	Trajectories	Joint SACo
Expert	12.77 ± 1.56	1.97	16.81	500000	1000	1.00

Environment name	Version	Agents	Action type	Observation size	Reward type
RWARE	Code included in Alberdice repository	4	Discrete	[71]	Dense

Converted from alberdice format to a Vault.

Uid	Episode return mean	Min return	Max return	Transitions	Trajectories	Joint SACo
Expert	15.67 ± 1.20	10.40	18.63	500000	1000	1.00

Environment name	Version	Agents	Action type	Observation size	Reward type
RWARE	Code included in Alberdice repository	6	Discrete	[71]	Dense

Converted from alberdice format to a Vault.

Uid	Episode return mean	Min return	Max return	Transitions	Trajectories	Joint SACo
Expert	17.45 ± 1.01	11.88	19.97	500000	1000	1.00