Speed and params question on MAPLE

[humphreyde]

HI, Thanks for your maple project!

I encountered some problems when I was trying train my own model and env

1.How to accelerate my training? Because I find maple is half as slow as benchmarking? Are there any settings or args that can speed up training?

2.Because I'm using my own robot and have changed the env(such as peg and hole's position) based on peg_ins, are there any settings or args in base_variant or env_params that I need to modify?

I would appreciate for your reply!!