Duplicate experiment will duplicate the current experiment as it was before you pushed the Run button and started the training.
To duplicate your model is a good alternative if you want to make bigger changes to your existing model.
Tips on how to improve an experiment
Other ways to continue experimenting
Tune experiment in the Evaluation view.
Use model adjustments that the platform suggests to see if you can potentially improve the performance.
When tuning, you copy the complete experiment without weights. That is, you do not keep what the model has learned previously.
Click Iterate in the Modeling view. This resumes training from a selected checkpoint with updated settings.
When iterating, you keep the weights. That is, you keep what the model has learned previously.
Use Continue training when an experiment showed really good progress up to a certain checkpoint but then it wasn’t so good anymore.
Reuse part of model lets you reuse a part of your model with weights from a specific checkpoint in a new model. For example, when you want to train a model with additional datasets.
How to duplicate an experiment
In the Modeling view, click Duplicate.
This will create a copy of your current model that you can edit, but the training progress will be lost.