Skip to content

Latest commit

 

History

History
22 lines (18 loc) · 533 Bytes

README.md

File metadata and controls

22 lines (18 loc) · 533 Bytes
Raw

ml_project_cuda

An implementation of Ojas learning rule written in cuda. The whole implementation will run on the GPU. Read the corresponding report for futher information

Requirements

  • Nvidia GPU supporting CUDA
  • Proprietairy Nvidia driver
  • CUDA toolkit (nvcc compiler)

Setup and running

  1. git clone
  2. cd to ml_project_cuda
  3. compile with nvcc using this command:
nvcc main.cu -rdc=true -o ojas
  1. run the code with profiler using this command:
nvprof ./ojas --profile-from-start off