Cannot use gpu to fine-tune on custom-data

[YkiWu]

I used "python main.py --base configs/custom_vqgan.yaml -t True --gpus 0" to fine-tune a pretrained model on my own data, but it seems that it always train on my cpu, not gpu, how should I solve this problem? Thanks! By the way, if I change the command into "python main.py --base configs/custom_vqgan.yaml -t True --gpus 0," just as says in README.md, I will got an error say that , error: argument --gpus: invalid int value: '0,'

[RichardXue123]

check your torch&torchvision if it is a cpu version

[jin1041]

use this python main.py --base configs/custom_vqgan.yaml -t True --gpus 0, add ',' at end