Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

my reproduction is so bad #12

Open
AIBUWAN opened this issue Sep 14, 2022 · 5 comments
Open

my reproduction is so bad #12

AIBUWAN opened this issue Sep 14, 2022 · 5 comments

Comments

@AIBUWAN
Copy link

AIBUWAN commented Sep 14, 2022

Hello @kwea123

I'm wondering why my reproduction is so bad, can you help explain it?

kid_reconstruction
2022-09-14 14-11-49 的屏幕截图

@kwea123
Copy link
Owner

kwea123 commented Sep 14, 2022

Do you get good result with the pretrained model?
What is your command to train?

@fuqichen1998
Copy link

I am having the same issue, could you help?

I am training on the jumping scene (12 frames). My command is

python train.py \
  --dataset_name monocular --root_dir /mnt/task_runtime/NSFF_Processed/jumping \
  --img_wh 960 540 --start_end 0 12 \
  --N_samples 128 --N_importance 0 --encode_t --use_viewdir \
  --num_epochs 50 --batch_size 512 \
  --optimizer adam --lr 5e-4 --lr_scheduler cosine \
  --exp_name jump_exp

@fuqichen1998
Copy link

I think there maybe some issue in the eval.py as the training samples in log are good.

@otonari726
Copy link

Do you add --use_viewdir option during the evaluation?
I had a similar problem, but adding the --use_viewdir option solved it.

@fuqichen1998
Copy link

Yes it works, thanks for the help!

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

No branches or pull requests

4 participants