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.

Sign up for GitHub

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

Jump to bottom

Questions about training procedure #5

Open

BrianWayland opened this issue Jan 19, 2022 · 0 comments

BrianWayland commented Jan 19, 2022

@Uehwan Thank you for sharing your implementation kindly. When learning your code I find two problems:

in train.py, the dataset is didn't looped through and just used the first batch for training, as shown in follows:

   ds = iter(trainer.train_loader)
   inputs = next(ds)

after predict depth and pose, the loss is not calculated to back propagate gradient and optimize parameters.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment