search

hikettei / cl-waffe2

[Experimental] Graph and Tensor Abstraction for Deep Learning all in Common Lisp
https://hikettei.github.io/cl-waffe2/
MIT License
122 stars 5 forks source link

[Optimize] 1.5x times faster training of MLP #96

Closed hikettei closed 10 months ago

hikettei commented 10 months ago

Changes

  1. CPUTensor is now able to use SLEEF.
  2. When loading training data, there's no need to make a copy
  3. Benchmarking MLP Training is now included in GNUmakefile