Define custom model metadata¶
To define metadata, create a model-metadata.yaml file and put it in the top level of the task/model directory. In most cases, it can be skipped, but it is required for custom transform tasks when a custom task outputs non-numeric data. The model-metadata.yaml is located in the same folder as custom.py.
The sections below show how to define metadata for custom models and tasks. For more information, you can review complete examples in the DRUM repository for custom models and tasks.
General metadata parameters¶
The following table describes options that are available to tasks and/or inference models. The parameters are required when using drum push to supply information about the model/task/version to create. Some of the parameters are also required outside of drum push for compatibility reasons.
Note
The modelID parameter adds a new version to a pre-existing custom model or task with the specified ID. Because of this, all options that configure a new base-level custom model or task are ignored when passed alongside this parameter. However, at this time, these parameters still must be included.
| Option | When required | Task or inference model | Description |
|---|---|---|---|
name |
Always | Both | A string, preferably unique for easy searching, that drum push uses as the custom model title. |
type |
Always | Both | A string, either training (for custom tasks) or inference (for custom inference models). |
environmentID |
Always | Both | A hash of the execution environment to use while running your custom model or task. You can find a list of available execution environments in Model Registry > Custom Model Workshop > Environments. Expand the environment and click on the Environment Info tab to view and copy the file ID. Required for drum push only. |
targetType |
Always | Both | A string indicating the type of target. Must be one of: • binary • regression • anomaly • unstructured (inference models only) • multiclass • textgeneration (inference models only) • transform (transform tasks only) |
modelID |
Optional | Both | After creating a model or task, it is best practice to use versioning to add code while iterating. To create a new version instead of a new model or task, use this field to link the custom model/task you created. The ID (hash) is available from the UI, via the URL of the custom model or task. Used with drum push only. |
description |
Optional | Both | A searchable field. If modelID is set, use the UI to change a model/task description. Used with drum push only. |
majorVersion |
Optional | Both | Specifies whether the model version you are creating should be a major (True, the default) or minor (False) version update. For example, if the previous model version is 2.3, a major version update would create version 3.0; a minor version update would create version 2.4. Used for drum push only. |
targetName |
Always | Model | In inferenceModel, a string indicating the column in your data that the model is predicting. |
positiveClassLabel / negativeClassLabel |
For binary classification models | Model | In inferenceModel, when your model predicts probability, the positiveClassLabel dictates what class the prediction corresponds to. |
predictionThreshold |
Optional (binary classification models only). | Model | In inferenceModel, the cutoff point between 0 and 1 that dictates which label will be chosen as the predicted label. |
trainOnProject |
Optional | Task | A hash with the ID of the project (PID) to train the model or version on. When using drum push to test and upload a custom estimator task, you have an option to train a single-task blueprint immediately after the estimator is successfully uploaded into DataRobot. The trainOnProject option specifies the project on which to train that blueprint. |
In the model-metadata.yaml file, you can also define runtime parameters to make your custom model code easier to reuse.