Rémi Louf
Rémi Louf
- [ ] Documentation - [ ] Sample (selected) deterministic variables - [ ] HMC from blackjax-devs/blackjax - [ ] NUTS from blackjax-devs/blackjax - [ ] Effective sample size from...
Opening this to have a discussion with myself about sampling UX.
Sometimes we'd like to help with an open source library but are not quite sure what with/don't want to bother the maintainer. I've been there. That is why I've a...
https://github.com/rlouf/mcx/blob/2a2b94801e68d94d86826863eeee80f0b84c390d/mcx/inference/hmc.py#L198-L205 We currently pass the parameters directly to the runtime; While the values of the parameters are passed in the `Trace` object, it would be convenient to update the Kernel's...
## Examples ˋmcx` is only interesting if it can be used and we need more example to showcase the way the library can be used. Examples are a great way...
Here are a few ideas for the core that I won't have the time to do before the first release but have the potential to make MCX more general and...
Tests not only allow to make sure the *code* is working as expected but they can also be used to make sure that the documentation and examples are up to...
`NetworkX`'s default is to separate the name of a node from the data attached to it; accessing a node by name `G.nodes["n"]` returns a dictionary with the data, here `{"contents":...
The following command ```python model[“variable_name”] ``` does not print anything interesting currently. Since every distribution implements “__repr__” we should print this at the very least.
It would be nice to be able to monitor the progress of sampling in real time. Totally do-able either using a callback and/or by logging every X steps when using...