Module For Hyperparamater Sweeps

Do you find it difficult to choose optimal hyperparameter values for training? What if you could experiment with a variety of hyperparameter values and their combinations? With this module, you can designate a range of values for a hyperparameter, specifying the step size for experimentation. Additionally, you have the flexibility to select combinations of various hyperparameters for your experiments.

How To Use This Module?

1. Install this module in your app where your dataset resides and your models will be stored
2. Open the installed module and select the model-type
3. Choose the correct dataset details and model template
4. Several hyperparamters would be dsiplayed once the template is chosen
5. To experiment on a hyperparamter, tick the Try a range of values checkbox and choose the min & max value for the range and specify the step size
6. Update the other hyperparamters values you wish to train on
7. To produce combinations of hyperparameter values, simply check the box labeled Generate Combinations
8. Select the combinations you want to train the model on and check the cost that will be incurred for each training on per hour basis
9. Once you click on Submit training jobs, model versions would be created on your Models section of the app
10. You can check for training & testing details on the Models page and see which model version performed the best