code related to imitation learning

[weixzhang]

Hello,

I hope this message finds you well. I've been exploring your repository and it seems that I couldn't locate the code related to imitation learning. Could you please guide me on how to utilize the data for imitation learning?

Thank you in advance for your assistance.

[yzqin]

Since this repo utilizes nearly the same file structure and environment codebase as DexArt, maybe you can take a look at the reinforcement learning/distillation guide here: https://github.com/Kami-code/dexart-release

[weixzhang]


Thank you so much for your prompt response and for pointing me in the right direction.
However, I apologize, but I seem to have encountered a difficulty. Despite following the reinforcement learning/distillation guide from the DexArt repository, I have not been able to locate the code specifically related to the DAPG algorithm within your repository.

Could you possibly provide further clarification or direct me to where I might find the DAPG-related code? Your additional support would be greatly appreciated as I continue to explore the potential of imitation learning with your data.

Thank you once again for your time and help.

[JackMa-coder]

Thank you so much for your prompt response and for pointing me in the right direction. However, I apologize, but I seem to have encountered a difficulty. Despite following the reinforcement learning/distillation guide from the DexArt repository, I have not been able to locate the code specifically related to the DAPG algorithm within your repository.  Could you possibly provide further clarification or direct me to where I might find the DAPG-related code? Your additional support would be greatly appreciated as I continue to explore the potential of imitation learning with your data.  Thank you once again for your time and help.

I also seem to be facing a challenge. After carefully following the reinforcement learning/distillation guide, I have not been able to locate the specific code for the DAPG algorithm within the repository.
Could you please assist me by pointing out where I might find the DAPG code? Any additional pointers or clarification would be immensely helpful as I am eager to proceed with my experiments.

If it’s not too much trouble, I would also appreciate any tips on how to navigate the repository or any other resources that might aid in my search.

Thank you once again for your time and assistance. I am looking forward to your response.

[yzqin]

In this work, we are using the vanilla DAPG, which is exactly the same as the one proposed in the original DAPG paper. I do not modify anything on the DAPG side and also the imitation learning algorithm is not the contribution of this work.

You may find our DAPG implementation with TRPO: https://github.com/yzqin/dexmv-learn/blob/master/mjrl/mjrl/algos/dapg.py
As well as the one implemented by the original author with NPG: https://github.com/aravindr93/mjrl/blob/master/mjrl/algos/dapg.py

[JackMa-coder]

In this work, we are using the vanilla DAPG, which is exactly the same as the one proposed in the original DAPG paper. I do not modify anything on the DAPG side and also the imitation learning algorithm is not the contribution of this work.

You may find our DAPG implementation with TRPO: https://github.com/yzqin/dexmv-learn/blob/master/mjrl/mjrl/algos/dapg.py As well as the one implemented by the original author with NPG: https://github.com/aravindr93/mjrl/blob/master/mjrl/algos/dapg.py

Thank you for your detailed response and for providing the relevant links.Your transparency and willingness to share resources are invaluable to the research community. I will definitely explore the provided implementations to further understand the work. Once again, thank you for your assistance!

[curieuxjy]

Hello @yzqin,
I’m trying to reproduce the training for this work and would appreciate it if you could provide the training entry-point script for the DAPG algorithm.

I found the DAPG-related code (for example, hand_teleop/dapg/dapg_wrapper.py) but couldn’t locate a ready-to-run training script or instructions on how to launch training. Could you either share the entry-point script or briefly describe how to run training with the existing files?

Thank you for your help.