search

DeMoriarty / fast_pytorch_kmeans

This is a pytorch implementation of k-means clustering algorithm
MIT License
284 stars 38 forks source link

if the data is [8,100000,64], 8 is batchs,Can it be processed in parallel? #1

Closed gcwang916 closed 3 years ago

gcwang916 commented 4 years ago

I have several sets of data here, but I want to process it in parallel, not use for i in range(): can you help me

DeMoriarty commented 4 years ago

@gcwang916 What do you mean by batch? do you want to train 8 seperate kmeans in parallel? if so you can try MultiKMeans. multi_kmeans.py. it can be used for product quantization if that's what you wanted.

DeMoriarty commented 3 years ago

closing due to no activity.