Skip to content

Faster Caffe Training

Domenic Curro edited this page Jun 8, 2016 · 4 revisions

Faster Caffe Training

Stanford University CS231n Lecture 11 gives useful tips for speeding up your training process:

  1. Use LMDB (seek times and no image decompression ie. jpeg).
  2. Use cuBLAS. This is a CUDA version of BLAS and will be faster than CPU optimized BLAS.
  3. Use cuDNN.
  4. Use 32bit floating point precision (when writing new layers), as they compute faster.