<!--
{
  "availability" : [
    "iOS: 15.0.0 -",
    "iPadOS: 15.0.0 -",
    "macCatalyst: 15.0.0 -",
    "macOS: 10.15.0 -",
    "tvOS: 16.0.0 -",
    "visionOS: 1.0.0 -"
  ],
  "documentType" : "symbol",
  "framework" : "CreateML",
  "identifier" : "/documentation/CreateML/MLSplitStrategy/automatic",
  "metadataVersion" : "0.1.0",
  "role" : "Case",
  "symbol" : {
    "kind" : "Case",
    "modules" : [
      "Create ML"
    ],
    "preciseIdentifier" : "s:8CreateML15MLSplitStrategyO9automaticyA2CmF"
  },
  "title" : "MLSplitStrategy.automatic"
}
-->

# MLSplitStrategy.automatic

Create ML automatically decides how much of your training dataset it uses for a validation dataset.

```
case automatic
```

## Discussion

Create ML creates a validation dataset by partitioning up to 10% from the training dataset, depending
on its size:

|Training samples|% used for validation|
|----------------|---------------------|
|< 50            |None                 |
|50 to 199       |10%                  |
|≥ 200           |5%                   |

---

Copyright &copy; 2026 Apple Inc. All rights reserved. | [Terms of Use](https://www.apple.com/legal/internet-services/terms/site.html) | [Privacy Policy](https://www.apple.com/privacy/privacy-policy)