热门标签
更多>
搜索结果
查询Tags标签: improvement,共有 2条记录-
Reinforcement Learning as One Big Sequence Modeling Problem
发表时间:2021 文章要点:这篇文章把RL看作序列建模问题(sequence modeling problem),直接用transformer来拟合整个序列(reats states, actions, and rewards as simply a stream of data,其实还拟合了reward-to-to return),拟合完了后就直接用这个transformer来做…
2021/8/28 6:06:07 人评论 次浏览 -
Reinforcement Learning as One Big Sequence Modeling Problem
发表时间:2021 文章要点:这篇文章把RL看作序列建模问题(sequence modeling problem),直接用transformer来拟合整个序列(reats states, actions, and rewards as simply a stream of data,其实还拟合了reward-to-to return),拟合完了后就直接用这个transformer来做…
2021/8/28 6:06:07 人评论 次浏览