search

WU-CVGL / BAD-NeRF

[CVPR 2023] 😈BAD-NeRF: Bundle Adjusted Deblur Neural Radiance Fields
https://wangpeng000.github.io/BAD-NeRF/
MIT License
182 stars 12 forks source link

Checkpoint Problem #11

Open Wang-Wilson opened 2 months ago

Wang-Wilson commented 2 months ago

Thanks for your great work! I am trying to do experiments on your code. I am not familiar with the code now, so I am inquiring about some bugs. I can load the weight you provided from the pre-trained model, but my PC accidentally shut down during the training. How can I start the training from my training checkpoint? I tried changing the weight_iter in configs to 220000, but it didn't work. However, when I change it back to 200000, It will start all over again.

The error is like:

Traceback (most recent call last): File "C:\Users\wilso\BAD-NeRF\train.py", line 324, in train() File "C:\Users\wilso\BAD-NeRF\train.py", line 109, in train graph.load_state_dict(graph_ckpt['graph']) File "C:\Users\wilso\anaconda3\lib\site-packages\torch\nn\modules\module.py", line 2153, in load_state_dict raise RuntimeError('Error(s) in loading state_dict for {}:\n\t{}'.format( RuntimeError: Error(s) in loading state_dict for Graph: Unexpected key(s) in state_dict: "se3_sharp.weight".

Do you have any idea about this?

wangpeng000 commented 1 month ago

Sorry to reply late, I re-download the weights I provided here weights and run the code successfully. Have you changed something? You can find the saved parameters in line 239.