droyed/eucl_dist

Euclidean Distance Computation in Python for 4x-100x+ speedups over SciPy and scikit-learn. Also leverages GPU for better performance on specific datasets.

PythonLGPL-3.0

Issues

error importing pycuda
#5 opened 4 years ago by AnyMail197
3
How did the author deal with Numerical stability problem?
#4 opened 5 years ago by liuhas2004
1
euclidean distance on 512 points
#3 opened 5 years ago by tomriddle54
1
In my test, scipy.dgemm is much slower(10x) than numpy.dot? Based on your test result, they seem to be close? Is it normal?
#1 opened 6 years ago by yxchng
1
Feature request: cosine distance
#2 opened 6 years ago by AruniRC
0