State Prediction on TD-MPC2 policy dataset Cheetah
3.8434MSEMoSim
Evaluation Results
| Method | Links | |
|---|---|---|
| MoSimTraining Data Source=random policy data, Horizon=1002025.04 | 3.8434 | |
| MoSimTraining Data Source=experience data, Horizon=162025.04 | 3.8841 | |
| DreamerV3Training Data Source=random policy data, Prediction Steps=5-step, Horizon=1002025.04 | 4.5623 | |
| DreamerV3Training Data Source=experience data, Prediction Steps=5-step, Horizon=162025.04 | 4.8908 | |
| MoSimTraining Data Source=random policy data, Horizon=162025.04 | 5.6052 | |
| DreamerV3Training Data Source=experience data, Prediction Steps=5-step, Horizon=1002025.04 | 9.427 | |
| DreamerV3Training Data Source=random policy data, Prediction Steps=5-step, Horizon=162025.04 | 11.1325 | |
| MoSimTraining Data Source=experience data, Horizon=1002025.04 | 13.0104 | |
| DreamerV3Training Data Source=experience data, Prediction Steps=1-step, Horizon=162025.04 | 13.1264 | |
| DreamerV3Training Data Source=experience data, Prediction Steps=1-step, Horizon=1002025.04 | 13.899 | |
| DreamerV3Training Data Source=random policy data, Prediction Steps=1-step, Horizon=162025.04 | 18.4404 | |
| DreamerV3Training Data Source=random policy data, Prediction Steps=1-step, Horizon=1002025.04 | 18.6197 |