Parallelize novel model fitting

finncatling / lap-risk

Uncertainty-aware mortality risk modelling in emergency laparotomy, using data from the NELA.

MIT License

5 stars 0 forks source link

Parallelize novel model fitting #70

Closed finncatling closed 3 years ago

finncatling commented 3 years ago

Each train-test split is taking 25 minutes on 8 cores of the main BDAU currently

finncatling commented 3 years ago

Parallelization is working on branch lap-risk-70 takes 26 minutes on 8 cores of the main BDAU. It is significantly slower still when run on all 32 cores (but they are much older CPUs) of the ganymede BDAU. I won't pursue parallelization any further