Browsing through the code, this looks a bit rough around the edges and untested. I'd wonder how it compares to similar packages for Spark, Hadoop or Flink. Lack of numbers and users make me wonder if it's anything more than someone's PhD project at CMU.
Comments
A comparison with http://prediction.io would be interesting.
Project lead Eric Xing has a nice talk about Petuum on Big Data Technology Conference 2014 in China: http://petuum.github.io/papers/eric_bdtc2014.pdf
Browsing through the code, this looks a bit rough around the edges and untested. I'd wonder how it compares to similar packages for Spark, Hadoop or Flink. Lack of numbers and users make me wonder if it's anything more than someone's PhD project at CMU.
I wonder how it compares to http://mahout.apache.org/
mahout is based on hadoop, mostly slower than cpp learning framework
http://paracel.io/
this is a similar project from douban inc.