train_own_dataset_tensorflow_colab
Tehsorflow colaboratory gave good chance to free use K80 GPU cards for trainings of own datasets. Max available time of training limited (12 hours) by session. For deep trainings only split trainings to steps.
Google Colab (Jupyter) notebook to retrain Object Detection Tensorflow model (ssd_mobilenet_v1_coco) with custom dataset (insulators).
Project idea based on RomRock story (https://hackernoon.com/object-detection-in-google-colab-with-custom-dataset-5a7bb2b0e97e) and Github
But prepared for last version of tensorflow detection models zoo(https://github.com/tensorflow/models/blob/master/research/object_detection/g3doc/detection_model_zoo.md) and for Python3.