Pytorch released version 1.0 in May, the “production ready pytorch”. “Married to Caffe2” and true production-readiness are still ideals on the horizon rather than present realities. Still, the idealogical shift points at a larger universal goal shared by the builders and users of ML infrastructure....
Context I spent the last couple of weeks learning how to optimize multithreading and multiprocessing to increase throughput during Tensorflow training. We have been able to overcome our CPU bottleneck, making our training GPU-limited. The library’s documentation is not inviting, and I found a lack of community information on the subject, so I thought I would summarize some of what I learned....
The Swamidass lab I am researching with has a tooling framework called Tflon. Matthew Mattlock, who wrote most of Tflon, first described the project as his attempt to provide tooling that shared Keras’s ease of use, but allowed for the full flexibiliy of Tensorflow....
I used the Hugo templating framework to build this site, and will list some of the pros/cons that I noticed. Hugo originally got my attention on HN, through various projects builing landing pages with the framework. Kubernetes also migrated their docs to Hugo recently, citing scaling issues with Jekyll....
I worked on a project awhile ago motivated by the inadequacies of the public pension system. We are approaching a tripling of retirement-age Americans. The pension systems are buckling due to poor market performance after the 2007 crash. Healthcare spending continues to rise ad neaseum....