pyro-ppl/pyro

[FR] Online (streaming) VI support

Open

#2,882 创建于 2021年6月21日

在 GitHub 查看
 (3 评论) (0 反应) (0 负责人)Python (8,211 star) (981 fork)batch import
enhancementhelp wanted

描述

It would be nice to have native support for online VI (its main 3 flavors are listed here: https://arxiv.org/abs/1904.03920 ).

Currently, the only tool applicable to the online learning scenario would be the SMCFilter, but it often performs poorly in high-dimensional situations.

SMCFilter itself could also benefit from particle MCMC rejuvenation as discussed here: https://www.springer.com/gp/book/9783030478445 and implemented in the particles package: https://github.com/nchopin/particles

贡献者指南

技术栈
pythonpytorch
领域
machine learning
议题类型
feature
难度面向新贡献者的预计实现难度，1 表示很小改动，5 表示专家级工作。
5
预计时间有经验贡献者完成调查、实现、测试并准备 pull request 的粗略时间范围。
over 1 week
活动状态议题当前的可参与程度：新鲜、活跃、陈旧、阻塞或等待维护者输入。
stale
清晰度议题是否清楚说明期望改动、验收标准和下一步。
needs investigation
前置要求
variational inferencePyro internalsSMCFilter
新手友好度1-100 的估计分数，表示该议题对首次贡献者的友好程度。
15
研究方向
Study the referenced arxiv paper (1904.03920) for online VI methods. Examine the current SMCFilter implementation in Pyro's source code. Review the particles package (nchopin/particles) for particle MCMC rejuvenation. Engage with maintainers to define the specific API and scope of the feature.