Tensorflow 2: Error when executing the command to train the model

I followed the Tensorflow 2 object detection tutorial by Gilbert Tanner: Tensorflow Object Detection with Tensorflow 2: Creating a custom model

Everything works fine until the last step when I have to train the model. Upon executing the following command: python model_main_tf2.py *
*** --pipeline_config_path=training/ssd_efficientdet_d0_512x512_coco17_tpu-8.config *
*** --model_dir=training ***
*** --alsologtostderr***

I get the following error:
Invoked with: None, 524288

Prior to this, I had an error which said: ValueError: numpy.ndarray size changed, may indicate binary incompatibility. Expected 88 from C header, got 80 from PyObject
However I was able to resolve this by changing the pycocotools version from 2.0.2 to 2.0.0.

Yet, I am unable to find a solution to the former error (Invoked with: None, 524288). I am executing this in Ubuntu 20.04, with python and pip versions of 3.8.10 and 21.2.4 respectively. The image attached contains a snapshot of the entire error. Kindly help me resolve this.


Welcome to the Tensorflow Forum!

Sorry for the late response.

Please check pipeline.config file parameters and ensure a valid value for each of the file parameters.

Thank you!