# Composable ML considerations

> Composable ML considerations - Platform support and considerations for working with Composable ML.

This Markdown file sits beside the HTML page at the same path (with a `.md` suffix). It summarizes the topic and lists links for tools and LLM context.

Companion generated at `2026-05-01T23:10:48.110685+00:00` (UTC).

## Primary page

- [Composable ML considerations](https://docs.datarobot.com/en/docs/reference/pred-ai-ref/cml-ref/cml-consider.html): Full documentation for this topic (HTML).

## Sections on this page

- [Environment support](https://docs.datarobot.com/en/docs/reference/pred-ai-ref/cml-ref/cml-consider.html#environment-support): In-page section heading.
- [DRUM supported OSs](https://docs.datarobot.com/en/docs/reference/pred-ai-ref/cml-ref/cml-consider.html#drum-supported-oss): In-page section heading.
- [Custom task languages](https://docs.datarobot.com/en/docs/reference/pred-ai-ref/cml-ref/cml-consider.html#custom-task-languages): In-page section heading.
- [Task limits](https://docs.datarobot.com/en/docs/reference/pred-ai-ref/cml-ref/cml-consider.html#task-limits): In-page section heading.
- [Modeling support](https://docs.datarobot.com/en/docs/reference/pred-ai-ref/cml-ref/cml-consider.html#modeling-support): In-page section heading.
- [Modeling specifics](https://docs.datarobot.com/en/docs/reference/pred-ai-ref/cml-ref/cml-consider.html#modeling-specifics): In-page section heading.
- [API](https://docs.datarobot.com/en/docs/reference/pred-ai-ref/cml-ref/cml-consider.html#api): In-page section heading.
- [Modeling options](https://docs.datarobot.com/en/docs/reference/pred-ai-ref/cml-ref/cml-consider.html#modeling-options): In-page section heading.
- [Insights and compliance documentation](https://docs.datarobot.com/en/docs/reference/pred-ai-ref/cml-ref/cml-consider.html#insights-and-compliance-documentation): In-page section heading.
- [Deployment options](https://docs.datarobot.com/en/docs/reference/pred-ai-ref/cml-ref/cml-consider.html#deployment-options): In-page section heading.

## Related documentation

- [Reference documentation](https://docs.datarobot.com/en/docs/reference/index.html): Linked from this page.
- [Predictive AI reference](https://docs.datarobot.com/en/docs/reference/pred-ai-ref/index.html): Linked from this page.
- [Composable ML reference](https://docs.datarobot.com/en/docs/reference/pred-ai-ref/cml-ref/index.html): Linked from this page.

## Documentation content

# Composable ML considerations

Consider the following when working with Composable ML.

### Environment support

Custom tasks are supported on DataRobot’s managed AI Platform (US, EU). Composable ML blueprints with DataRobot tasks are available for all users.

#### DRUM supported OSs

- DRUM works on Mac OS, Linux, and Windows 10. Testing is on Linux only and therefore compatibility issues may occur when running on other platforms.

#### Custom task languages

- Python and R are supported.
- SAS is not supported.

#### Task limits

| Component | Size |
| --- | --- |
| RAM | 60GB training / 14GB scoring |
| CPU Cores | 4 |
| GPU | not supported |
| Storage | 350GB |
| Artifact (Serialized trained model/transformer size) | 10GB |
| Timeout | 72 hrs (for fit) |
| Max custom tasks per blueprint | 3 |

### Modeling support

The following describes modeling-specific capabilities.

#### Modeling specifics

The following are supported:

- Predictive ML, including time-aware but not time series, and Feature Discovery.
- Estimators, both built-in and custom, are available for binary classification, regression, multiclass, anomaly detection, and clustering.
- Preprocessing, both built-in and custom.

#### API

- Python API client for blueprint generation.
- Python client for custom task generation.

#### Modeling options

The following describes support for advanced modeling options with custom blueprints:

- Metrics and loss function, where “loss function” is used to train a model and “metric” is used to evaluate models and for accuracy monitoring. Typically, the same measure is used for both. Support includes:
- Exposure, Weight, Counts of Events, Offset NoteDataRobot does not indicate whether a task takes into account any of these options. As a result, using Composable ML on a project that uses those options is not recommended. (For example, if you train a custom blueprint on a project that uses exposure, there is no guarantee whether the model will useyory/exposureas the target.) On the other hand, when using blueprints generated by Autopilot, these options are taken into account correctly, in line with the project settings.
- Monotonic constraints are not supported.
- Hyperparameter tuning is supported for custom and built-in tasks.
- Blenders are supported with the exception of custom estimators. You can howevermanuallycreate a blueprint that uses multiple estimators.

#### Insights and compliance documentation

- Model-agnosticinsights are supported:
- Model-specificinsights offer limited support:
- Compliance documentation is supported.

### Deployment options

- All MLOps model monitoring and management features are supported.
- Deployment of a custom blueprint inside the DataRobot platform is supported.
- Deployment outside of the DataRobot platform (using Scoring Code or Portable Prediction Server) is only supported for blueprints without custom tasks.