Locomotion on D4RL HalfCheetah-medium v2
98.7Online Return (Normalized)BOORL
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| BOORLevaluation_rollouts=10, step_interval=50002026.05 | 98.7 | 47.9 | |
| ROIDseeds=3, evaluation_rollouts=10, step_interval=50002026.05 | 95.51 | 47.48 | |
| RLPDevaluation_rollouts=10, step_interval=50002026.05 | 90.5 | — | |
| Cal-QLevaluation_rollouts=10, step_interval=50002026.05 | 72.3 | 48 | |
| PEXevaluation_rollouts=10, step_interval=50002026.05 | 67.8 | 47.3 | |
| DACERTime steps=1M, Number of random seeds=52026.05 | 49.92 | — | |
| DUALTime steps=1M, Number of random seeds=52026.05 | 49.83 | — | |
| EDISTime steps=1M, Number of random seeds=52026.05 | 49.34 | — | |
| Diff-QLTime steps=1M, Number of random seeds=52026.05 | 48.61 | — | |
| IQLTime steps=1M, Number of random seeds=52026.05 | 48.35 | — | |
| DiffuserTime steps=1M, Number of random seeds=52026.05 | 48.2 | — | |
| ODTevaluation_rollouts=10, step_interval=50002026.05 | 42.1 | 42.7 |