4 ms·
The authors report speedups of a factor of 3-4 by experimenting with BLAS and GPU-based convolutional neural network implementations in C++.
by henning 17y ago
The authors report speedups of a factor of 3-4 by experimenting with BLAS and GPU-based convolutional neural network implementations in C++.