How to reproduce the Rainbow results of your paper

[jiawei415]

Hi. I see that the result about Rainbow in your paper is not the same as [1][2].
So, I wanted to ask what hyperparameter settings were used for the Rainbow results in your paper.
Hope you can give a detailed introduction on how to use the code in README, such as the requirements, thanks!

[1] Van Hasselt H P, Hessel M, Aslanides J. When to use parametric models in reinforcement learning?[J]. Advances in Neural Information Processing Systems, 2019, 32.
[2] Kaiser Ł, Babaeizadeh M, Miłos P, et al. Model Based Reinforcement Learning for Atari[C]//International Conference on Learning Representations. 2019.