Hi all, I am using a super computer with a 48-hour time-limit, and I have four different time points (four jobs), only one of which finishes on time. The others return with nothing; which is a sign that the job was too long. I am wondering how I can work around this…maybe to set the four jobs to on…

Time Limit on Super Computer; How to Finish from Where Iterations Left Off?

mike-lawrence January 4, 2021, 4:50pm 13

Note that there seems to be a bug in achieving this; see discussion here.

Topic		Replies	Views
Stan - do several runs of estimation Modeling	2	543	February 10, 2022
How to improve model speed as the number of datapoints increase Modeling fitting-issues , performance	7	374	September 17, 2021
Sampler getting stuck? Modeling	18	3314	May 26, 2019
Model taking too long Algorithms rstan , fitting-issues	1	376	November 4, 2022
Parallelizing Bayesian Hierarchical Model with Many Parameters Modeling performance , paralellization	6	468	October 27, 2023