Developer guide: Create a model run via the Python SDK
A model run is a model training experiment within a model directory. Each model run provides a versioned snapshot of the data rows, annotations, and data splits for that model run. You can upload predictions to the model run and compare its performance against other model runs in a model directory.
Once you create a model and choose the labeled data rows for training, you can add a model run to keep track of your model training experiment. It is important to put your model run into the model directory with the correct data rows and ontology configured in the previous steps.
If you are iterating multiple model experiments on a machine learning task, the best practice is to put your model runs under the same model directory. This allows you to visualize and compare the performance of the different model runs.
During the model run creation process, you can specify how you want to split the selected data rows into training, validation, and testing sets. To learn more, see Splits.
Labelbox versions your data rows, annotations, and data splits at the time of training, making it easy for you to reproduce your model and compare models trained on different data versions.
First, select the appropriate model from the Model tab. Make sure to choose the right model directory, because the model run will access the data rows and ontology defined by the model creation step. If you already have model runs in this model directory, you should see the previous model runs' evaluation results.
To create a new model run, click New model run. This will lead you to configure a new model experiment and create data splits.
Provide a name for the model run.
Select how to construct your data split. You can opt to Load data splits from previous run if previous model runs have been created in this model directory, or you can choose Set data split to customize the values.
Click Create model run to complete the process.
Click the settings icon in the top right corner of the model runs page to modify the selected model run. To change the selected model run, select an option from the leftmost dropdown. You can update a model run in the following ways:
- Rename model run to a different name.
- Update the Model run config (for more details, see the documentation here).
Additionally, you can Copy model run ID from this menu of options.
To delete a model run, select the desired model run from the leftmost dropdown. Then click the settings icon in the top right corner of the model runs page and choose Delete model run. You will need to confirm the deletion before proceeding.
Updated 5 months ago