Cannot get results from pb model.

[OctavianM13]

[Required] Your device (RaspberryPi3, LaptopPC, or other device name): Laptop with Mac OS

[Required] Your device's CPU architecture (armv7l, x86_64, or other architecture name): x86_64

[Required] Your OS (Raspbian, Ubuntu1604, or other os name): Mac OS

[Required] Details of the work you did before the problem occurred:

I trained a one class classifier using darknet tiny yolo from here: https://github.com/AlexeyAB/darknet and got .weights file from the training.

I converted .weights file to .pb using convert_weights_pb.py.

I run demo.py with these options: --input_img ./0b86cecf-67d1-4fc0-87c9-b36b0ee228bb.jpg --output_img ./res.jpg --data_format NHWC --frozen_model ./frozen_darknet_yolov3_model.pb --class_names yolov3-tiny-mine.names.
[Required] Error message:


2019-10-23 10:38:30.254189: I tensorflow/core/platform/cpu_feature_guard.cc:141] Your CPU supports instructions that this TensorFlow binary was not compiled to use: AVX2 FMA
Loaded graph in 0.25s
Predictions found in 1.89s

This is the output of the prediction


[Required] Overview of problems and questions:

When I test .weights I get results for my input, but after the conversion the output si the same as the input - empty (with no bounding boxes).