In this repository, we try to solve musculoskeletal tasks with `Double DQN reinforcement learning` by using a `transformer` model has been used as the base model architecture.
Why do you think that https://github.com/wandb/wandb is a good alternative to MyosuiteDDQN