Reinforcement Learning on DMC1000K difficult tasks vector inputs (train test)
863Quadruped-RunA2-winner-v
Evaluation Results
| Method | Links | ||||||
|---|---|---|---|---|---|---|---|
| A2-winner-vInput Type=vector, Number of Seeds=10, Evolution Search Domain=vector-based tasks2022.10 | 863 | 213 | 200 | 329 | 311 | 75 | |
| A2-winnerInput Type=vector, Number of Seeds=10, Evolution Search Domain=image-based tasks2022.10 | 838 | 278 | 579 | 286 | 299 | 88 | |
| SACInput Type=vector, Number of Seeds=102022.10 | 707 | 134 | 379 | 7 | 2 | 1 | |
| CURLInput Type=vector, Number of Seeds=102022.10 | 497 | 60 | 363 | 7 | 2 | 1 | |
| SAC-IdentityInput Type=vector, Number of Seeds=102022.10 | 345 | 121 | 506 | 9 | 16 | 1 |