Hello Everyone,
I have recently started working with Isaac Lab which is a very deep and complex topic involving frameworks outside of the Isaac Sim/Lab remit.
I have noticed that many people on the forums and discord are struggling with where to start in training and deploying their own models. The devs have done a great job recently improving documentation and example code for ease of use. Even while this is the case, the wider topic of training and deploying models is something that even more seasoned programmers and devs are struggling with.
Ultimately, I would say the majority of us want to use some kind of AI in robotic control or detection.
I personally feel that while there is always room for Nvidia to provide better tools, workflows and documentation to make it easier for us, it’s also a “You problem” as they can’t give us every use case that we might want to achieve in the using the software and developing our own solutions.
There is currently an Open USD study group in the forum that meets once a week on a Zoom call. I wanted to float the idea that us as the community get together and share information to improve our understanding of the wider world of Isaac Lab workflows and RL methods.
I’m just floating the idea and wondering if there is interest to make this happen?
If you are interested, please drop a comment or drop me a PM if you think you would like to be involved.
Thanks
Scott