Decanter AI handles different algorithms with different hyperparameter combinations. Therefore, every algorithm is capable of training multiple models.

You can specify the number of models to train in the [Builder] tab under [Advanced Settings]. By specifying the number of models to train, Decanter AI will limit the number of models trained during an experiment. Typically we do this to reduce the model training time, the more models Decanter AI trains the longer it takes to finish the model training process.

Depending on the plan you're on, the maximum number of models to train can be different.

1. If you are on the Individual plan, the maximum number of models for an experiment is 50

2. If you are on the On-Premise plan, the maximum number of models for an experiment is 200