ggmm
ggmm copied to clipboard
ebattenberg
Python module to train GMMs using CUDA
Will need to adapt sklearn full covariance routines to CUDAMat.
- [x] Simplify method naming to be more clear (originally used sklearn method names) - [ ] Make sure public method calls return numpy array (not CUDAMat arrays) - [...
According to the mccabe static analyzer, some methods have too many independent paths through the code and need to be refactored. These methods were taken from sklearn and and ported...