Hi,
Thank you so much for your work, it inspires me a lot. One thing I'm not clear is whether I should use action sampling in the rollout phase. From the PPO blog's code (repo: CleanRL, The 37 Implementation Details of Proximal Policy Optimization), seems it should have. I hope you can help me explain. Thank you
Hi, Thank you so much for your work, it inspires me a lot. One thing I'm not clear is whether I should use action sampling in the rollout phase. From the PPO blog's code (repo: CleanRL, The 37 Implementation Details of Proximal Policy Optimization), seems it should have. I hope you can help me explain. Thank you
Best regards, Yichao