Repository navigation
|
Thanks for the project! As the topic said, I used 0.12mm and a 15% infill as the default printing parameters, but I assume there are still some subtle weight differences that will be caused by using different 3D print machines or differenrt 3d print materials. So assume that the weight of my printing parts will be slightly different from yours; will this subtle difference make the sim2real fail? |
Answered by
MarcDcls
Aug 8, 2026
Replies: 1 comment
|
Technically, it does impact the sim-to-real gap, but the RL policy should be robust enough to handle such small weight differences. The policy was trained with domain randomization, which specifically helps compensate for this kind of discrepancy. As long as you don't deviate too much from the recommended print parameters, you should be fine 😉 |
0 replies
Answer selected by
MarcDcls
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Technically, it does impact the sim-to-real gap, but the RL policy should be robust enough to handle such small weight differences. The policy was trained with domain randomization, which specifically helps compensate for this kind of discrepancy. As long as you don't deviate too much from the recommended print parameters, you should be fine 😉