Falkon v0.6

Implementation of an in-core version of Falkon, suitable for smaller data analyses. Here the data is always kept inside the GPU, thus the model can train much faster. The result is the InCoreFalkon class.

LAUUM was improved to use a CUDA implementation for the inner-loop function.

API changes are limited to the FalkonOptions class, where certain option names have changed.

Several bug fixes were also introduced, and edge-cases fixed.