search

GeorgeSeif / Semantic-Segmentation-Suite

Semantic Segmentation Suite in TensorFlow. Implement, train, and test new Semantic Segmentation models easily!
2.51k stars 879 forks source link

Exception due to variable reuse #76

Closed Nosferath closed 6 years ago

Nosferath commented 6 years ago

I am running the default custom model, using my own dataset. I already trained the model, and I am trying to perform the test using arguments --dataset crack_images_v2 --crop_height 96 --crop_width 96 --model custom --mode test --continue_training True. However, the following Exception appears:

`ValueError: Variable logits/weights already exists, disallowed. Did you mean to set reuse=True or reuse=tf.AUTO_REUSE in VarScope? Originally defined at:

File "/home/inti/anaconda3/envs/TF18-GPU/lib/python3.6/site-packages/tensorflow/contrib/framework/python/ops/variables.py", line 246, in variable use_resource=use_resource) File "/home/inti/anaconda3/envs/TF18-GPU/lib/python3.6/site-packages/tensorflow/contrib/framework/python/ops/arg_scope.py", line 183, in func_with_args return func(*args, **current_args) File "/home/inti/anaconda3/envs/TF18-GPU/lib/python3.6/site-packages/tensorflow/contrib/framework/python/ops/variables.py", line 291, in model_variable use_resource=use_resource)`

I cannot find a variable named logits/weights in order to enable reuse. I am having the same problem if I ever try to continue training with the default model, and with RefineNet (all variants). How do I get around this?

Nosferath commented 6 years ago

Nevermind, solved by restarting the python kernel.