zju3dv / ENeRF

SIGGRAPH Asia 2022: Code for "Efficient Neural Radiance Fields for Interactive Free-viewpoint Video"
https://zju3dv.github.io/enerf
Other
418 stars 28 forks source link

KeyError: 'rgb_level0' #8

Closed wjlyu closed 1 year ago

wjlyu commented 1 year ago

I retrain with zjumocap dataset with command: python train_net.py --cfg_file configs/enerf/zjumocap_eval.yaml while I get the error: Setting up [LPIPS] perceptual loss: trunk [vgg], v[0.1], spatial [off] Loading model from: /opt/conda/lib/python3.8/site-packages/lpips/weights/v0.1/vgg.pth Traceback (most recent call last): File "train_net.py", line 117, in main() File "train_net.py", line 109, in main train(cfg, network) File "train_net.py", line 51, in train trainer.train(epoch, train_loader, optimizer, recorder) File "/dfs/data/ENeRF/lib/train/trainers/trainer.py", line 56, in train output, loss, loss_stats, image_stats = self.network(batch) File "/opt/conda/lib/python3.8/site-packages/torch/nn/modules/module.py", line 1015, in _call_impl return forward_call(*input, **kwargs) File "lib/train/losses/enerf.py", line 23, in forward color_loss = self.colorcrit(batch[f'rgb{i}'], output[f'rgb_level{i}']) KeyError: 'rgb_level0'

wjlyu commented 1 year ago

I have fix the problem, because of level0 is not rendered. thanks for your great work!