beanmachine
beanmachine copied to clipboard
A library that allows for inference on probabilistic models
PLEASE GO HERE INSTEAD: https://github.com/facebookresearch/beanmachine/pull/1581
Summary: Added a stanza in the pybinding file to enable call to collect_statistics from Python code. Extended the file graph_test.py with a relevant test case. Differential Revision: D38088017
Summary: This is a very simple diff that 1) binds the statistics method in BMG into python 2) tests that it works. This enables python code to access the statistics...
### Motivation This PR introduces LLVM, MLIR, and PyBind11 as dependencies to the Beanmachine. I will add the demo code as a separate PR once this is merged. ### Changes...
Add llvm
### Motivation Adds llvm/mlir dependency as git submodule. A git submodule offers us the most flexibility for the following reasons: - MLIR is not tied to a version of LLVM;...
Summary: This is a demonstration of compiling a Python function to MLIR to LLVM IR to machine code then executing the program with the input from python. It is effectively...
Summary: With the move of functions from MH to Graph, there was a name collision of functions both called `compute_support`, but with one of them returning node ids and the...
### Motivation This is the PR where all implementations of MixedHMC live along with surrounding docs and tutorial. This should not be merged yet. ### Types of changes - [...
Reviewed By: yucenli Differential Revision: D37051297
Summary: This test case is a simplified (that is, k=2) version of the GEP model. Reviewed By: mootaz77 Differential Revision: D34314625