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
Precondition Error when is_training is set to false #17
Comments
This is due to a bug in the training code as the |
Hi, Have you tried training with moving mean/variance restored while testing? I tried save everything while training, and restore bn parameters when testing, but got much worse results. |
What were your batchnorm hyperparameters? The tensorflow default 'decay' for batchnorm (https://www.tensorflow.org/api_docs/python/tf/contrib/layers/batch_norm) seems too high from my preliminary experiments. I will update the code with proper batchnorm configuration soon (most likely within a week). |
@tinghuiz Thank you for the response. I have used the default slim parameters (same as in your code). With default setting, the decay is 0.999. |
From some online discussion of the batch_norm layer, decay of 0.999 is not desirable for relatively small-scale problems (i.e. problems that don't require millions of training steps). Can you try a smaller decay such as 0.9 or 0.95 and see if it helps? |
I will try. One more question about bn is that for train_op, only trainable_variables are fed into the optimizer. My reading from the documents is that bn parameters are not in the trainable_variable list but in global_variable list. Is the bn mean and variance changing if train_op only applies on trainable_vars? |
Good point. You should replace it with something like |
I have removed batch_norm altogether in the latest update. |
how to save moving_mean and moving_variebles to trainable_variables ? have solved ? |
I noticed that when the depth test graph is being build, the
is_training
argument fordisp_net
is not set to False. Won't this negatively affect the test performance, as the batch normalization won't be configured properly?When setting this argument to True, an exception is raised. (Related to batch norm)
I get this when using the model that was provided in the "download_model.sh" script
The text was updated successfully, but these errors were encountered: