Fabian
The lower bound for the random reset space of the object in disassemble is higher than the upper bound for the first index. This appears to be an issue for...
Currently, the observation space does not specify a dtype, which means it defaults to float32, however the returned numpy array is by default in float64. This will lead to exceptions...
Recently, other big benchmark suites, such as [OpenAI gym](https://github.com/openai/gym/blob/024b0f5160716b01effc27de45615bac7620e8b6/README.md?plain=1#L51-L52) and [DMC](https://github.com/deepmind/dm_control/blob/41d0c7383153f9ca6c12f8e865ef5e73a98759bd/README.md?plain=1#L61-L68), switched to the new [pybind11-based mujoco bindings](https://pypi.org/project/mujoco/). Since these bindings are made available in DeepMind's official mujoco repository, I...
I was trying to use your environments as standalone gym tasks without the metalearning aspect. While I do not think this issue conflicts with the design choices made for the...
Currently the render function of MujocoEnv does not take a `mode` argument for the `render()` function and instead exposes the `offscreen` flag. https://github.com/rlworkgroup/metaworld/blob/a0009ed9a208ff9864a5c1368c04c273bb20dd06/metaworld/envs/mujoco/mujoco_env.py#L121-L133 This is an issue when using the...
Currently, only some of the tasks are described into more detail. This should be extended to all available tasks.
Currently, the render function only supports "human" format and does not return anything for the array case. This should be improved to used Black-Box envs with vision obervations.
Quite some ppl are using the [hand environments](https://github.com/aravindr93/hand_dapg), it might be interesting to add the to the repo as well.