Simple explanation about ODHUB training parameters

This section aims to provide simple instructions on how to select the right training parameter based on the project's requirement and your training engine.

Model type

ODHUB provides three different model types to represent the inference speed and accuracy of the trained model

• FAST will reduce model accuracy in exchange for inference speed
• ACCURATE is the balance between inference speed and accuracy
• MOST ACCURATE will reduce inference speed in exchange for model accuracy
• CUSTOM is an under-development feature that will allow users to add more model types for their training.

Choosing the suitable model type will depend on your application's requirements, but the accurate and most accurate type will take more time for the training process unless you have a robust GPU.

Batch size

Divide the training dataset into smaller groups to reduce GPU/CPU load.

Number of steps

During each step, an amount of batch size is processed. The more number of steps, the better the trained model is.

For Example: during one step, assuming that batch size is 20, it means that 20 images will be processed at the same time.

Learning rate

The default value is recommended.