search

intel / ideep

Intel® Optimization for Chainer*, a Chainer module providing numpy like API and DNN acceleration using MKL-DNN.
MIT License
165 stars 90 forks source link

modify convolution_forward to support broadcast add fusion #337

Closed jiayisunx closed 2 weeks ago

jiayisunx commented 2 weeks ago

@jgong5, could you please review this PR? Thanks!