object detection api - Tensorflow- 'No Variable To Save'

Question

I am busy following this tutorial and I have completed all the steps barring training my model

When I run the command to train I receive several of the the following errors:

• W0128 17:45:44.627660 5284 variables_helper.py:155] Variable [ssd_mobile_net_v1fpn_keras_feature_extractor/conv_pw_4_bn/moving_variance] is not available in checkpoint

Which culminate in the following error - "ValueError: No variables to save". I have now tried this with two models and it is causing the same error. I am on a windows machine, and my train config is set to include from_detection_checkpoint: true and fine_tune_checkpoint_type: "detection".

Thank you so much

question from:https://stackoverflow.com/questions/65940578/tensorflow-no-variable-to-save

Answer 1

Waitting for answers