search

VITA-Group / EnlightenGAN

[IEEE TIP] "EnlightenGAN: Deep Light Enhancement without Paired Supervision" by Yifan Jiang, Xinyu Gong, Ding Liu, Yu Cheng, Chen Fang, Xiaohui Shen, Jianchao Yang, Pan Zhou, Zhangyang Wang
Other
918 stars 200 forks source link

Problems about how to fine-tuning #125

Closed p994070003 closed 2 years ago

p994070003 commented 2 years ago

Hello Sir.

Thanks for releasing this amazing implementation, the results are great!

I found no the code about how to fine tune the modle. I wanna continue to train the model with my dark datasets in trainA and the dataset that you provided in trainB by a smaller learning rate,How should I operate it?

Any help will be appreciated!

Thanks. Nova.

jiejiangd commented 2 years ago

想请问一下这个可以单卡训练吗

yifanjiang19 commented 2 years ago

finetuning is not supported, you might need to change the code a little bit. E.g., loading the pretrained weight before starting your training job. @p994070003

yifanjiang19 commented 2 years ago

@jiejiangd single-gpu is doable but not guaranteed to reproduce it