I’ve a flashed Mac 5,1 operating 10.13.6, and a Geforce GTX 760 GPU with the most recent supported CUDA library put in at a compute functionality 3. Mathematica 11 makes use of the GPU fairly properly for deep studying (6X speedup effectively outpacing my newer MacMini) however I would like to have a look at utilizing Python or comparable ideally utilizing TensorFlow, Keras or Pytorch, since Mthematica 11’s DL assist is not nice. Putting in the most recent variations of theses libraries clearly is just not going to work since they now want a better compute functionality.
Are there any archived variations of those libraries or different deep studying instruments that may set up efficiently?
If no such can be found wouldn’t it be value upgrading from a 16G 4 core field to a 128Gb 12 core field, dropping the GPU and run TensorFlow or such in core solely?
