Support different codebook sizes in CUDA kernels

[dryman]

Does this package support other dimensions than 128?
We'd like to test it on GloVe, which is 100.

[una-dinosauria]

Yep, just add your data reading functions to

Rayuela.jl/src/read_datasets.jl

Line 4 in ccf22ab

function read_dataset(

[dryman]

Does the following code in cuda matter?

Rayuela.jl/deps/src/cudautils.cu

Lines 93 to 95 in ccf22ab

	// FIXME hard-coding 256 entries in each codebook, and 128 dimensions
	const int H = 256; // size of each codebook
	const int D = 128; // dimensionality of each vector

[una-dinosauria]

Oh right, sorry. The GPU part is hardcoded.

[una-dinosauria]

I checked again, since I definitely ran this code on the GPU for my thesis.

The function veccost that you linked above is not called from the Julia part. veccost2 is

Rayuela.jl/deps/src/cudautils.cu

Lines 154 to 155 in ccf22ab

	// FIXME hard-coding 256 entries in each codebook
	const int H = 256; // size of each codebook

which only hardcodes the number of entries in the codebook (not the dataset dimensionality). I'm changing the name of this issue to reflect that.

[una-dinosauria]

Also opened #41 to keep track of code deletion.