issues Search Results · repo:NVlabs/HOVER language:Python
Filter by
22 results
(67 ms)22 results
inNVlabs/HOVER (press backspace or delete to remove)when I use this repo to train teacher policy on unitree G1，I find my teacher policy can not cpnverging . My retargeting
data is from huggingface . I used the complete motion set from this paper
zzl19950704
- Opened 10 days ago
- #33
In the function below, I am not sure about the shape of the matrice: def compute_joint_tracking_error( joint_pos:
torch.Tensor, joint_pos_gt: torch.Tensor, frame_weights: torch.Tensor, num_envs: int ) ...
attila-lingenieur
- Opened 23 days ago
- #30
Hi @pulkitg01,
While delving deeper into your code, I noticed that you re assigning the projected_gravity value for the student policy
using the function asset.data.projected_gravity_b. From what I understand, ...
longdvt
- 2
- Opened 25 days ago
- #29
Hi, I m a bit confused about this line here. Can you explain why you use mean * 0.0 + std here? Wouldn t it be std not
changing throughout the whole training
process?https://github.com/NVlabs/HOVER/blob/43ee08b637c84f3ad11b6af45a89586bf7f65630/third_party/rsl_rl/rsl_rl/modules/actor_critic.py#L129 ...
percypeng5221
- Opened 28 days ago
- #28
Hi, I m running
${ISAACLAB_PATH:?}/isaaclab.sh -p scripts/rsl_rl/play.py \
--num_envs 10 \
--reference_motion_path neural_wbc/data/data/motions/amass_all_ACCAD.pkl \
--student_player \
...
- 1
- Opened on May 3
- #27
I used the stable_punch.pkl in h2o, but after running it, I found an error indicating that the dagger data was
missing.Please help me.
Traceback (most recent call last): File /home/chenwenhao/chenwenhao/HOVER/scripts/rsl_rl/train_teacher_policy.py ...
VinHo-C
- 1
- Opened on Apr 28
- #25
I built my own robot use this respository. but when I train the robot, it looks that feet can not contact groud in
IsaacSim. so where to set the robot height when train the policy? in mujoco all things ...
rcwind
- 1
- Opened on Apr 27
- #24
Hello,
I was reviewing your implementation of reference trajectories, and I have a few questions.
In this issue, you mentioned that training for 80k iterations results in a performant policy. I m curious ...
- 2
- Opened on Apr 25
- #23
Thank you for the great work! Do you have a plan to support Unitree G1?
Or
Could you please provide some suggestions about extending HOVER to support G1?
cjianli
- Opened on Apr 16
- #22
Hello, I try to train the student policy using distill_mask_modes = { omnih2o : DISTILL_MASK_MODES_ALL[ omnih2o ]}. The
teacher policy is trained for 120000 iterations and student policy is trained for ...
ZuhongLIU
- 4
- Opened on Apr 14
- #21
Learn how you can use GitHub Issues to plan and track your work.
Save views for sprints, backlogs, teams, or releases. Rank, sort, and filter issues to suit the occasion. The possibilities are endless.Learn more about GitHub IssuesProTip!
Press the /
key to activate the search input again and adjust your query.Learn how you can use GitHub Issues to plan and track your work.
Save views for sprints, backlogs, teams, or releases. Rank, sort, and filter issues to suit the occasion. The possibilities are endless.Learn more about GitHub IssuesProTip!
Restrict your search to the title by using the in:title qualifier.