ManiSkill2-Learn
ManiSkill2-Learn copied to clipboard
haosulab
Hello, I noticed that in the implementation of SAC, when using shared_backbone=True the visual_nn of the actor is used in the critic_cfg. I understand the intention of sharing the visual...
I store trajectories generated with my RL agent in a h5 file using the functionality from the [evaluation.py ](https://github.com/haosulab/ManiSkill2-Learn/blob/main/maniskill2_learn/env/evaluation.py) script. Now I want to train a network on those trajectories....
Hi, I am trying to train an agent with dapg+ppo on the turn faucet environment using rgbd observations. I have used the demonstration conversion script to convert the demonstrations in...
The __init__.py file of the "modules" package tries to import the pn2_modules file. However, that script does not exist. I would like to ask whether there are implementations of Pointnet++...
Hello, I want to train a DAPG+PPO agent using 5 GPUs. I want to use the same hyperparameters as specified in the ManiSkill2 paper in Appendix F.8 in Table 9:...
I would like to ask what are the operating parameters for the point cloud of BC's pickcube, such as using OBS_ Is the frame ee or base? I saw that...
Is there any example or tutorial about deploying a trained agent (e.g., BC agent with control mode "pd_ee_delta_pose") on the real-world Franka robot arm? For the sim2real experiments in the...
I found for some environments like PickSingleYCB, the evaluation protocol in the documents shows that there are not only 5 episodes per object in the training set (74 YCB objects)...
### Compile Version information ``` mani-skill2 0.4.2 maniskill2-learn 1.8.0b0 warp-lang 0.4.3 ``` Compile command ```python -m warp_maniskill.build_lib --cuda_path /usr/local/cuda-11.3 --mode debug``` Result: ``` Building /home/rl/anaconda3/envs/mani38/lib/python3.8/site-packages/warp_maniskill/warp/bin/warp.so g++ -O0 -g -D_DEBUG -DWP_CPU...
