autograd mir and CUDA library for dynamic neural networks in D.
Variable(T, size_t dim, alias Storage)
unlike numpy$ dub --config=example-mnist -b=cuda-release # with cuda
$ dub --config=example-mnist -b=release # without cuda
it results as following (may take several seconds without cuda)
Running ./grain-example-mnist
loading data/train-images-idx3-ubyte.gz
loading data/train-labels-idx1-ubyte.gz
loading data/t10k-images-idx3-ubyte.gz
loading data/t10k-labels-idx1-ubyte.gz
train loss: 0.538635, acc: 0.864311
test loss: 0.299959, acc: 0.915264
train loss: 0.277901, acc: 0.920858
test loss: 0.241783, acc: 0.930589
train loss: 0.229879, acc: 0.934999
test loss: 0.206087, acc: 0.939704
train loss: 0.198716, acc: 0.943937
test loss: 0.181938, acc: 0.945613
train loss: 0.175066, acc: 0.950957
test loss: 0.163919, acc: 0.951022
$ curl -fsS https://dlang.org/install.sh | bash -s ldc-1.9.0
$ source ~/dlang/ldc-1.9.0/activate
$ dub test -b=cuda-unittest # with cuda
$ dub test # without cuda
I have tested with
CUDA in D
Referenced autograd libraries
sorted by higher priority for me
dub --config=example-mnist
dub --config=example-char-rnn