add support for half precision gemm

[bjarthur]

in conjunction with FluxML/NNlib.jl#363, add support for half-precision gemm, for which a special kernel is provided by Nvidia. see JuliaGPU/CUDA.jl#1080

[mcabbott]

Why do you say this is needed in addition? It looks like an alternative path. But the existing method NNlib._batched_gemm!(::Type{<:CuArray}, ought to match Float16 (if NNlib.jl would let it be called).

What would be good to add here is tests using this precision. Which I think should test the user-facing batched_mul not the internal functions.

[DhairyaLGandhi]

Why would nnlib prevent it from getting called?

[bjarthur]

the current code actually works with Float16, but falls back to batched_mul_generic! where a loop is performed over the last dimension. so painfully slow. i thought about tests, but couldn't come up with a way to test that the batched nvidia kernel is called instead.

[ToucheSir]

Yup, the overriden method in NNlib uses BlasFloat, which does not include Float16. Now, one hang-up I see with this PR is that _batched_try_gemm! also only accepts BlasFloat. @bjarthur can you confirm this works locally without any errors?

[bjarthur]

indeed, it does work locally without any errors, otherwise i would not have submitted it! ;)

[ToucheSir]

Great, I think per @mcabbott's comment a test for this would be good :)