Code

Build example 3 if CUDA enabled#222

Code

Merged

jwallwork23 merged 15 commits intomainCambridge-ICCS/FTorch:mainfrom 208_multi-gpu-buildCambridge-ICCS/FTorch:208_multi-gpu-buildCopy head branch name to clipboardJan 29, 2025

+100-41

Commits

Commits on Jan 29, 2025

• Build example 3 if CUDA and MPI enabled

  

  jwallwork23

  authored and

  jatkinson1000

  committed
  a763b0e

  Copy full SHA for a763b0e

• Put model on CUDA device in simplenet

  

  jwallwork23

  authored and

  jatkinson1000

  committed
  eadc273

  Copy full SHA for eadc273

• Run example 3 if it's been built

  

  jwallwork23

  authored and

  jatkinson1000

  committed
  e16c17d

  Copy full SHA for e16c17d

• Add missing imports for pt2ts

  

  jwallwork23

  authored and

  jatkinson1000

  committed
  1392a0a

  Copy full SHA for 1392a0a

• More helpful output for simplenet_infer_python

  

  jwallwork23

  authored and

  jatkinson1000

  committed
  0129030

  Copy full SHA for 0129030

• Fix numbering in CMakeLists for examples

  

  jwallwork23

  authored and

  jatkinson1000

  committed
  2392649

  Copy full SHA for 2392649

• Renaming in MultiGPU example; set up unit testing

  

  jwallwork23

  authored and

  jatkinson1000

  committed
  6a532b5

  Copy full SHA for 6a532b5

• Raise error if no CUDA in example 3

  

  jwallwork23

  authored and

  jatkinson1000

  committed
  e8b5780

  Copy full SHA for e8b5780

• Lint

  

  jwallwork23

  authored and

  jatkinson1000

  committed
  37d5da8

  Copy full SHA for 37d5da8

• Fix model filename passed to fortran

  

  jwallwork23

  authored and

  jatkinson1000

  committed
  bec0734

  Copy full SHA for bec0734

• Do require mpi4py in Python script

  

  jwallwork23

  authored and

  jatkinson1000

  committed
  ef0f938

  Copy full SHA for ef0f938

• Drop ENABLE_MPI CMake argument

  

  jwallwork23

  authored and

  jatkinson1000

  committed
  e0d4e45

  Copy full SHA for e0d4e45

• Bugfix: Update example 3 README with correct names of files simplenet -> multigpu.

  jatkinson1000

  committed
  8bcb9d0

  Copy full SHA for 8bcb9d0

• Bugfix: explicitly move loaded net to required device in multigpu python

  Show description for a1f5688
  jatkinson1000

  committed
  a1f5688

  Copy full SHA for a1f5688

• Bugfix: Include missing CheckLanguage module in the CMakeLists of example 3.

  jatkinson1000

  committed
  ef07fca

  Copy full SHA for ef07fca