error Found array with 0 sample(s) (shape=(0, 512)) while a minimum of 1 is required

[AlexeyAL1e]

i am trying to train a neural network on my own dataset and have it classify the test dataset with high accuracy
Firstly, I formed my dataset of 30 personalities, each person has 34 face photos. after that i used #356 (comment) commands to align and get the loss with using script softmax.py
Second, I had to run this script to get the .pb file
python src/freeze_graph.py .model_dir output_file.pb
Thirdly I got a .pkl file using this
python src/classifier.py TRAIN ~/datasets/my_dataset/train/ ~/models/model-20170216-091149.pb ~/models/my_classifier.pkl --batch_size 34
Now, I use my model to identify faces in my test dataset, with run script:
python src/classifier.py CLASSIFY ~/datasets/my_dataset/test/ ~/models/model-20170216-091149.pb ~/models/my_classifier.pkl --batch_size 34
but I get an error Found array with 0 sample(s) (shape=(0, 512)) while a minimum of 1 is required, it looks like the file .pkl, there are no entries for embedded images
Tell me, if these are all points, then what am I doing wrong?

[Zhou-Yicheng]

python src/classifier.py CLASSIFY ~/datasets/my_dataset/test/ ~/models/model-20170216-091149.pb ~/models/my_classifier.pkl --batch_size 34

Change to
python src/classifier.py CLASSIFY ~/datasets/my_dataset/train/ ~/models/model-20170216-091149.pb ~/models/my_classifier.pkl --batch_size 34 ?