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

Dataset comments #7

Open
Kjos opened this issue Mar 22, 2018 · 1 comment
Open

Dataset comments #7

Kjos opened this issue Mar 22, 2018 · 1 comment

Comments

@Kjos
Copy link

Kjos commented Mar 22, 2018

Hi,
Thanks for putting this up on github. It got me started with tensorflow.
However some comments;
I think the dataset shows a lot of underfitting. I get a validation score of 0.5 at end, or even below.
I connected tensorboard and the learning graph shows classic underfitting signs (straight learning rate).
Perhaps it doesn't help that the images have wide ranging aspect ratios?
I tried to insert horizontal flipped images as well, but it doesn't help that much.

@sardarwasif
Copy link

Hello,
I am training neural network using this code (cv-tricks.com-master\cv-tricks.com-master\Tensorflow-tutorials\tutorial-2-image-classifier\train) on my own data set of road traffic signs in which i have left and right road signs. I have changed the directory as follows:

classes = os.listdir("D:\My_tensorflow\data_set\Train")
train_path="D:\My_tensorflow\data_set\Train"
test_path="D:\My_tensorflow\data_set\Test"

when i run this code it gives this kind of out put and error:
Number of files in Training-set: 0
Number of files in Validation-set: 0
error is:
assert batch_size <= self._num_examples

Kindly help me to figure out this problem.
Thanks

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

2 participants