ml4a / ml4a

A python library and collection of notebooks for making art with machine learning.

Home Page:https://ml4a.net

Geek Repo:Geek Repo

Github PK Tool:Github PK Tool

running PCA crashes my computer

rosenb0rg opened this issue · comments

When I try to run PCA in the eigenfaces.ipynb notebook my computer crashes (cold boot). When I watch the system monitor it appears that the CPU is working at over 100% so it seems the problem would be there. I'm running Debian Stretch with an Intel i7-4790K CPU and a GeForce GTX 970 GPU. Here is the specific line of code:

n_components = 100
pca = PCA(n_components=n_components, svd_solver='randomized', whiten=True)
pca.fit(X)

I've tried reducing the components to 50 to no avail.

Any advice on how to diagnose the problem would be greatly appreciated.

update: I went through the whole rigamarole of getting my CUDA drivers properly installed and getting tensorflow-gpu into my anaconda environment. The issue seems to be that PCA just overloads my CPU and it shuts down my computer, and I can't direct it to use my GPU