YingfanWang/PaCMAP

Implementation with PyTorch

Opened this issue · 1 comments

Even with CPU thread limiting applied while using the provided code, the CPU consumption remains excessively high (around 1400%). Is it possible to implement this algorithm in PyTorch?

I have implemented a basic version of the algorithm in Javascript using Tensorflow.js You can probably convert it to PyTorch using that as a reference https://github.com/zqwitt/PaCMAP_TFjs