Dataset Cards - OMIGA
2c_vs_64zg - Download
Metadata
| Environment name | Version | Agents | Action type | Observation size | Reward type |
|---|---|---|---|---|---|
| SMAC (v1) | Modified version of SMAC v1, popularised by MAPPO | 2 | Discrete | [478] | Dense |
Generation procedure for each dataset
Converted from omiga format to a Vault.
Summary statistics
| Uid | Episode return mean | Min return | Max return | Transitions | Trajectories | Joint SACo |
|---|---|---|---|---|---|---|
| Poor | 8.91 ± 1.01 | 2.53 | 10.00 | 10830 | 348 | 1.00 |
| Medium | 13.00 ± 1.39 | 10.01 | 15.00 | 37940 | 1001 | 1.00 |
| Good | 19.94 ± 1.26 | 15.18 | 21.61 | 59215 | 1001 | 1.00 |
6h_vs_8z - Download
Metadata
| Environment name | Version | Agents | Action type | Observation size | Reward type |
|---|---|---|---|---|---|
| SMAC (v1) | Modified version of SMAC v1, popularised by MAPPO | 6 | Discrete | [172] | Dense |
Generation procedure for each dataset
Converted from omiga format to a Vault.
Summary statistics
| Uid | Episode return mean | Min return | Max return | Transitions | Trajectories | Joint SACo |
|---|---|---|---|---|---|---|
| Poor | 9.12 ± 0.81 | 4.80 | 9.99 | 24255 | 1001 | 1.00 |
| Medium | 11.97 ± 1.26 | 10.00 | 14.99 | 29511 | 1001 | 1.00 |
| Good | 17.84 ± 2.15 | 15.01 | 20.02 | 38040 | 1001 | 1.00 |
5m_vs_6m - Download
Metadata
| Environment name | Version | Agents | Action type | Observation size | Reward type |
|---|---|---|---|---|---|
| SMAC (v1) | Modified version of SMAC v1, popularised by MAPPO | 5 | Discrete | [124] | Dense |
Generation procedure for each dataset
Converted from omiga format to a Vault.
Summary statistics
| Uid | Episode return mean | Min return | Max return | Transitions | Trajectories | Joint SACo |
|---|---|---|---|---|---|---|
| Poor | 8.50 ± 1.19 | 1.81 | 9.89 | 22747 | 1001 | 0.96 |
| Medium | 11.03 ± 0.58 | 10.08 | 11.96 | 27717 | 1001 | 0.95 |
| Good | 20.00 ± 0.00 | 20.00 | 20.00 | 27734 | 1001 | 0.96 |
corridor - Download
Metadata
| Environment name | Version | Agents | Action type | Observation size | Reward type |
|---|---|---|---|---|---|
| SMAC (v1) | Modified version of SMAC v1, popularised by MAPPO | 6 | Discrete | [346] | Dense |
Generation procedure for each dataset
Converted from omiga format to a Vault.
Summary statistics
| Uid | Episode return mean | Min return | Max return | Transitions | Trajectories | Joint SACo |
|---|---|---|---|---|---|---|
| Poor | 4.93 ± 1.71 | 0.00 | 9.99 | 51268 | 1001 | 1.00 |
| Medium | 13.07 ± 1.27 | 10.02 | 14.99 | 126012 | 1001 | 1.00 |
| Good | 19.88 ± 1.01 | 15.01 | 20.49 | 100170 | 1001 | 1.00 |
6halfcheetah - Download
Metadata
| Environment name | Version | Agents | Action type | Observation size | Reward type |
|---|---|---|---|---|---|
| MAMuJoCo | V1.0, Mujoco v200 | 6 | Continuous | [23] | Dense |
Generation procedure for each dataset
Converted from omiga format to a Vault.
Summary statistics
| Uid | Episode return mean | Min return | Max return | Transitions | Trajectories | Joint SACo |
|---|---|---|---|---|---|---|
| Medium-Replay | 655.76 ± 590.40 | -198.77 | 2132.60 | 1001000 | 1000 | 1.00 |
| Medium-Expert | 2105.38 ± 1073.24 | 251.94 | 3866.09 | 2002000 | 2000 | 1.00 |
| Medium | 1425.66 ± 520.12 | 251.94 | 2113.52 | 1001000 | 1000 | 1.00 |
| Expert | 2785.10 ± 1053.14 | 317.94 | 3866.09 | 1001000 | 1000 | 1.00 |
2ant - Download
Metadata
| Environment name | Version | Agents | Action type | Observation size | Reward type |
|---|---|---|---|---|---|
| MAMuJoCo | V1.0, Mujoco v200 | 2 | Continuous | [113] | Dense |
Generation procedure for each dataset
Converted from omiga format to a Vault.
Summary statistics
| Uid | Episode return mean | Min return | Max return | Transitions | Trajectories | Joint SACo |
|---|---|---|---|---|---|---|
| Medium-Replay | 1029.51 ± 141.27 | 895.37 | 1517.06 | 1751750 | 1750 | 0.66 |
| Medium-Expert | 1736.88 ± 319.64 | 840.77 | 2124.15 | 2002000 | 2000 | 1.00 |
| Medium | 1418.70 ± 37.04 | 840.77 | 1473.86 | 1001000 | 1000 | 1.00 |
| Expert | 2055.07 ± 22.07 | 1994.03 | 2124.15 | 1001000 | 1000 | 1.00 |
3hopper - Download
Metadata
| Environment name | Version | Agents | Action type | Observation size | Reward type |
|---|---|---|---|---|---|
| MAMuJoCo | V1.0, Mujoco v200 | 3 | Continuous | [14] | Dense |
Generation procedure for each dataset
Converted from omiga format to a Vault.
Summary statistics
| Uid | Episode return mean | Min return | Max return | Transitions | Trajectories | Joint SACo |
|---|---|---|---|---|---|---|
| Medium-Replay | 746.42 ± 671.89 | 70.76 | 2801.15 | 1314826 | 4160 | 1.00 |
| Medium-Expert | 1190.61 ± 973.40 | 95.27 | 3762.69 | 1919782 | 5481 | 1.00 |
| Medium | 723.57 ± 211.66 | 128.38 | 2776.49 | 919391 | 4000 | 1.00 |
| Expert | 2452.02 ± 1097.86 | 95.27 | 3762.69 | 1000391 | 1481 | 1.00 |