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
Can we store the fine-tuned weight using tf.train.Saver()? #21
Comments
No, you can use this code and store variables using tensorflow saver. This is infact done in the KittiSeg repository. Take a closer look at demo.py, where a trained KittiSeg model, which is based on fcn8, is loaded. I am using .npy to store the original vgg weights, as the .ckpt format is very unflexable. You cannot easily add, remove or rename variabels after storing the graph to .ckpt. This is however very useful to use this code as part of more advanced models. For you purpose saving to .ckpt should be fine. |
@MarvinTeichmann Thank you for your quick reply! I checked the KittiSeg, but all I can found is stuff like |
Checkpoints are loaded in |
Hi, I am fine-tuning your FCN toward my own dataset. But after tuning, I was doing stuff like
'conv1_1/filter:0','conv1_1/biases:0'; sess.run(filters)
to extract features and save them into npy file. But since we have lots of filters here, listing all of them would be messy.So I was wondering if we can do it with the tf.train.Saver, which allows us to store the trained Variables fast and clean. But later I notice that there is no "Variable" in the fcn8s_vgg file. And I am thinking that by using your original fcn8s_vgg.py file, we can't use the tf.train.Saver(), right?
I am new to tensorflow, please let me know your comments, thanks.
The text was updated successfully, but these errors were encountered: