Segmentation fault while training

google / hypernerf

Code for "HyperNeRF: A Higher-Dimensional Representation for Topologically Varying Neural Radiance Fields".

https://hypernerf.github.io

Apache License 2.0

895 stars 105 forks source link

Segmentation fault while training #16

Open dolcelee opened 2 years ago

dolcelee commented 2 years ago

I got this segmentation fault when I start to train the HyperNeRF. I thought it was an OOM error, but it still shows up after I reduced the batch_size. I reinstalled numpy as well, nothing happened. Any ideas would be appreciated. f59d7298ddeeffa6d4a2dc96e15a6fc

MasatoKawai commented 2 years ago

how about modify requirements.txt to

numpy==1.19.3 #(change numpy version) jaxlib==0.1.71 #(delete +cuda111)