Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Add an example on particle filtering / particle Gibbs #43

Open
yebai opened this issue Aug 1, 2024 · 4 comments
Open

Add an example on particle filtering / particle Gibbs #43

yebai opened this issue Aug 1, 2024 · 4 comments

Comments

@yebai
Copy link
Member

yebai commented Aug 1, 2024

It would be good to have a simple example of particle filtering. The previous version depends on AdvancedPS; we could remove that dependency to keep this minimal.

https://github.com/TuringLang/SSMProblems.jl/tree/8a458bd2941d38e75932683f9dc994d5dd80c47c/examples/smc

@FredericWantiez
Copy link
Member

I don't think it depends on AdvancedPS but happy to rewrite it with the new interface

@yebai
Copy link
Member Author

yebai commented Aug 1, 2024

That sounds good.

A fun experiment: You can even try vectorising the particle sampling process to see whether it works well with Metal.jl.

@THargreaves
Copy link
Collaborator

Just a heads up about Metal.jl. I ran a few benchmarks with it on my M1 Mac and found that even for easily parallelisable functions such as sum(), it ran the same speed as on the CPU.

Still an interesting experiment but don't be surprised if it feels slow.

@yebai
Copy link
Member Author

yebai commented Aug 1, 2024

M-chips seem to have good support for vectorisation, so I won't be surprised that sum works comparably on CPU and Metal. Do you benchmark anything more substantial?

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

3 participants