search

autonomousvision / monosdf

[NeurIPS'22] MonoSDF: Exploring Monocular Geometric Cues for Neural Implicit Surface Reconstruction
MIT License
572 stars 53 forks source link

The final output result cannot be found #82

Open wujinhu1999 opened 1 year ago

wujinhu1999 commented 1 year ago

I'm not using distributed training, I changed the code slightly, the command I run on the terminal is:python training/exp_runner.py --local_rank=2 --conf confs/dtu_mlp_3views.conf --scan_id 65，and then： image After a period of training, the program stops without outputting any additional information: image Looking forward to your reply.

niujinshuchong commented 1 year ago

At your first training, the cuda kernel needs to be complied and it might take several minutes.

wujinhu1999 commented 1 year ago

But I have trained several times and still haven't seen the result.

niujinshuchong commented 1 year ago

Dose it start training?

wujinhu1999 commented 1 year ago

I want to ask if it has started training? But it does not output training process information. image

niujinshuchong commented 1 year ago

I think no as it will print the loss if it starts training.