Published January 28, 2021
| Version 1.3.1
Software
Open
ReactiveMP.jl: a Julia package for automatic Bayesian inference on a factor graph with reactive message passing.
Description
ReactiveMP v3.0.0
Closed issues:
- Networks of discrete random variables (#98)
- Test rules macro fails with indexed rules (#216)
- Fix broken demo on Kalman Smoothing with neural network driven dynamics (#225)
- Free energy calculation with mean-field constrain encounters a MethodError (#233)
Merged pull requests:
- Add CVI implementation for the delta node (#188) (@Nimrais)
- Add
getnodefnfunction in rules (#217) (@bvdmitri) - Support indexed rules in
@call_rulemacro (#218) (@bvdmitri) - Develop delta fn eus fix indexed call rule (#219) (@bvdmitri)
- Develop delta fn cvi fix indexed rules (#220) (@bvdmitri)
- Fix interaction between
autovarand variables in a form of a tuple (#221) (@bvdmitri) - Fix KLSTM demo (#227) (@Chengfeng-Jia)
- Reduce the training time of neural network (#228) (@Chengfeng-Jia)
Notes
Files
biaslab/ReactiveMP.jl-v3.0.0.zip
Files
(520.8 kB)
| Name | Size | Download all |
|---|---|---|
|
md5:a8c7214ea774955c2a3a48e399c4402c
|
520.8 kB | Preview Download |
Additional details
Related works
- Is supplement to
- https://github.com/biaslab/ReactiveMP.jl/tree/v3.0.0 (URL)