We read every piece of feedback, and take your input very seriously.
To see all available qualifiers, see our documentation.
This is my personal CNN forward implementation. Using CUDA and OpenCL as its backends.
