Step by Step walk through of model playground experiments.

After creating the splits, the user can now run experiments. By clicking a split, the user is able to create an experiment.

In the experiment, the user is able to specify the name of the experiment and has the option to use the Hasty template for their model.

On choosing the hasty template, the architectures, its parameter values, loss metrics, and all other features of a model will be automatically set for that specific machine learning task. For example for instance segmentation task, the Basic Instance Segmentor template would be used.

If the users want to design their own model with the provided architectures, solvers, optimizers, loss metrics, etc., then that is also possible by not using the template.

To learn more about the different aspects of the model:

Model Aspects Resource
Architecture architectures
Augmentations augmentations
Scheduler schedulers
Loss metrics loss
Solver solvers
Training Parameters training-parameters
Other Metrics metrics

Here, the user can find a detailed explanation of all the model parameters provided by hasty.

Boost model performance quickly with AI-powered labeling and 100% QA.

Learn more
Last modified