Open mcfefa opened 5 years ago
There's no such constraint in the algorithm itself, but you may be hitting e.g. the maximum memory per user constraint set by your OS. To better diagnose it we would need the output from Julia, in particular the stack trace.
Has anyone run into issues with running TSne.jl on a cluster and returning a "Segmentation fault" error? I've run a few different datasets with TSne and some appear to be running fine on the cluster, but two keep throwing a Segmentation fault error while attempting to run the code. The two that continue to fail seem to be the largest datasets. Are there any data constraints I need to be concerned about when running TSne?