cyy/go2_rl_robotlab
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

update better default training settings; update new policy and results; update readme.

Browse Source
This commit is contained in:
wertyuilife2
2026-06-29 03:08:09 +08:00
parent f08058fcb8
commit 28b4516d22
9 changed files with 15 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files

19
README.md
View File

@@ -73,12 +73,12 @@ It is an official [MoE-CTS](https://robogauge.github.io/static/files/arxiv.pdf)
<img src="resources/results/robogauge_compare.png" width="100%"/>
</p>
### Algorithm Results (Best of 150k training steps)
### Algorithm Results (Best Checkpoint Results)
| Model | Score | Tracking | Safety | Quality | Level |
| --- | --- | --- | --- | --- | --- |
| go2_moe_cts (go2_rl_robotlab) | **0.6828** | **0.6785** | 0.7552 | **0.7645** | **8.17** |
| go2_moe_cts (go2_rl_gym) | **0.6713** | 0.6669 | **0.7857** | 0.7392 | 7.85 |
| go2_moe_cts (go2_rl_robotlab) | **0.6984** | **0.7055** | **0.8159** | **0.7693** | **8.30** |
| go2_moe_cts (go2_rl_gym) | 0.6713 | 0.6669 | 0.7857 | 0.7392 | 7.85 |
| [CTS](https://arxiv.org/pdf/2405.10830) vanilla | 0.5786 | 0.5755 | 0.7066 | 0.6624 | 6.83 |
| [HIM](https://github.com/InternRobotics/HIMLoco) | 0.5379 | 0.5453 | 0.6476 | 0.6050 | 6.19 |
| [DreamWaQ](https://arxiv.org/abs/2301.10602) | 0.5054 | 0.5105 | 0.6149 | 0.5730 | 5.74 |
@@ -251,16 +251,17 @@ xml_path: "{ROOT_DIR}/resources/go2/your-custom-scene.xml"
## Differences from `go2_rl_gym`
- Motor:
- use official unitree motor model instead of simple PD controller
- use official unitree motor model instead of simple PD controller.
- Rewards:
- different tracking reward form (fixed sigma vs. dynamic sigma)
- lower joint_acc_l2 weight in Lab due to physics-step level implementation and sensitivity to outliers
- extra joint_pos_penalty_l1 reward in Lab due to better performance
- 2x tracking reward weights due to better performance.
- different tracking reward form (fixed sigma vs. dynamic sigma).
- lower joint_acc_l2 weight in Lab due to physics-step level implementation and sensitivity to outliers.
- extra joint_pos_penalty_l1 reward in Lab due to better performance.
- Domain randomization:
- no randomized action delay, use motor-level delay instead
- no randomized action delay, use motor-level delay instead.
- no motor strength randomization due to implementation constraints in Lab.
- History length: 10 in Lab vs. 5 in Gym, due to better performance with longer history in Lab.
- Terrain Difficulty: use lab's continuous difficulty levels.
---
## Acknowledgements