Zheng Xiong

Support for robots with different DoFs in Mujoco XLA

4

Hi, Thanks for developing the amazing mujoco v3 with JAX support! I'm working on a scenario where I want to train a single RL policy on multiple robots with different...

Any way to get expert demos for each task?

3

Hi, Thanks for open-sourcing this brilliant benchmark! I was wondering that is there any quick way to get expert demos for each task in Simpler? Like a script oracle policy,...

How to convert RT-1-X action to Octo action

3

Hi, I wanted to use successful episodes collected by RT-1-X in the google robot envs as demos to fine-tune an Octo model. I was wondering that how could I convert...

How to specify which objects to manipulate in pick and move tasks

5

Hi, I was wondering how can I specify which object(s) to manipulate in the `pick objects` and `move A to B` tasks? Based on my understanding of the code, the...

Hi, I'm using [parallel env](https://github.com/Lifelong-Robot-Learning/LIBERO/blob/f78abd68ee283de9f9be3c8f7e2a9ad60246e95c/libero/libero/envs/venv.py#L924) to evaluate on a LIBERO task with different initial states. I use 20 parallel envs and it takes about 30 seconds to close them by...

Generating LoRA parameters with hypernetworks

3

Hi, Thanks for open-sourcing this brilliant package! Similar to #6 , I also want to apply hypernetworks (HN) to LoRA, but my setting is a bit different. I have a...

Hi, Thanks for open-sourcing this great dataset! I'm trying to download the bridge V2 dataset from [https://rail.eecs.berkeley.edu/datasets/bridge_release/data/](url), but experiences very slow download speed of ~100kb/s. I'm from University of Oxford...