search

openai / glow

Code for reproducing results in "Glow: Generative Flow with Invertible 1x1 Convolutions"
https://arxiv.org/abs/1807.03039
MIT License
3.11k stars 515 forks source link

Finetune weights #80

Open anth2o opened 5 years ago

anth2o commented 5 years ago

Hi,

I was wondering if there was a way to use the weights trained on the CelebA generation task and to fine tune them on a custom dataset. I already tried to use the weights loaded from 'graph_optimized.pb' with the restore_path arguments but it doesn't work. Does someone have those weights usable in a training configuration, or knows a way to load them correctly ?

Thanks in advance