Closed arjun370 closed 4 years ago
In the last sprint, we compared hyper-parameters influencing the accuracy of the SSD mobilenet v2 network.
The goal is to freeze the configuration parameters for future simulations.
e. As a baseline, the final model configuration file and annotated data which achieved the best overall result (softmax) for the projects are included in the tensorflow_training folder of the repo.
In the last sprint, we compared hyper-parameters influencing the accuracy of the SSD mobilenet v2 network.
The goal is to freeze the configuration parameters for future simulations.