Ultralytics HUB Integrations

Learn about Ultralytics HUB integrations with various platforms and formats to streamline your AI workflows.

Datasets

Seamlessly import your datasets into Ultralytics HUB for efficient model training.

Once a dataset is imported, you can train a model on it just as you would with native Ultralytics HUB datasets.

Roboflow

You can easily filter Roboflow datasets on the Ultralytics HUB Datasets page.

Ultralytics HUB screenshot of the Datasets page with Roboflow provider filter

Ultralytics HUB supports two types of integrations with Roboflow: Universe and Workspace.

Universe

The Roboflow Universe integration allows you to import one dataset at a time into Ultralytics HUB from Roboflow.

Import

When exporting a Roboflow dataset, select the Ultralytics HUB format. This action redirects you to Ultralytics HUB and opens the Dataset Import dialog.

Import your Roboflow dataset by clicking the Import button.

Ultralytics HUB screenshot of the Dataset Import dialog with an arrow pointing to the Import button

Next, you can train a model on your newly imported dataset.

Ultralytics HUB screenshot of the Dataset page of a Roboflow Universe dataset with an arrow pointing to the Train Model button

Remove

Navigate to the Dataset page of the Roboflow dataset you wish to remove. Open the dataset actions dropdown menu and click the Remove option.

Ultralytics HUB screenshot of the Dataset page of a Roboflow Universe dataset with an arrow pointing to the Remove option

Tip

You can also remove an imported Roboflow dataset directly from the main Datasets page.

Ultralytics HUB screenshot of the Datasets page with an arrow pointing to the Remove option of one of the Roboflow Universe datasets

Workspace

The Roboflow Workspace integration allows you to import an entire Roboflow Workspace at once into Ultralytics HUB.

Import

Navigate to the Integrations page by clicking the Integrations button in the sidebar.

Enter your Roboflow Workspace private API key and click the Add button.

Tip

Clicking the Get my API key button will redirect you to your Roboflow Workspace settings, where you can find your private API key.

Ultralytics HUB screenshot of the Integrations page with an arrow pointing to the Integrations button in the sidebar and one to the Add button

This connects your Ultralytics HUB account with your Roboflow Workspace, making your Roboflow datasets available within Ultralytics HUB.

Ultralytics HUB screenshot of the Integrations page with an arrow pointing to one of the connected workspaces

Next, you can train a model using any of the datasets from the connected workspace.

Ultralytics HUB screenshot of the Dataset page of a Roboflow Workspace dataset with an arrow pointing to the Train Model button

Remove

Navigate to the Integrations page via the sidebar. Click the Unlink button for the Roboflow Workspace you want to disconnect.

Ultralytics HUB screenshot of the Integrations page with an arrow pointing to the Integrations button in the sidebar and one to the Unlink button of one of the connected workspaces

Tip

You can also unlink a connected Roboflow Workspace directly from the Dataset page of any dataset belonging to that workspace.

Ultralytics HUB screenshot of the Dataset page of a Roboflow Workspace dataset with an arrow pointing to the remove option

Tip

Alternatively, remove a connected Roboflow Workspace directly from the main Datasets page using the remove option associated with any dataset from that workspace.

Ultralytics HUB screenshot of the Datasets page with an arrow pointing to the Remove option of one of the Roboflow Workspace datasets

Models

Exports

After you train a model, you can export it to 13 different formats using the Export mode, including popular ones like ONNX, OpenVINO, CoreML, TensorFlow, and PaddlePaddle.

Ultralytics HUB screenshot of the Deploy tab inside the Model page with an arrow pointing to the Export card and all formats exported

The available export formats are detailed in the table below.

Format	`format` Argument	Model	Metadata	Arguments
PyTorch	-	`yolo11n.pt`	✅	-
TorchScript	`torchscript`	`yolo11n.torchscript`	✅	`imgsz`, `optimize`, `nms`, `batch`
ONNX	`onnx`	`yolo11n.onnx`	✅	`imgsz`, `half`, `dynamic`, `simplify`, `opset`, `nms`, `batch`
OpenVINO	`openvino`	`yolo11n_openvino_model/`	✅	`imgsz`, `half`, `dynamic`, `int8`, `nms`, `batch`, `data`
TensorRT	`engine`	`yolo11n.engine`	✅	`imgsz`, `half`, `dynamic`, `simplify`, `workspace`, `int8`, `nms`, `batch`, `data`
CoreML	`coreml`	`yolo11n.mlpackage`	✅	`imgsz`, `half`, `int8`, `nms`, `batch`
TF SavedModel	`saved_model`	`yolo11n_saved_model/`	✅	`imgsz`, `keras`, `int8`, `nms`, `batch`
TF GraphDef	`pb`	`yolo11n.pb`	❌	`imgsz`, `batch`
TF Lite	`tflite`	`yolo11n.tflite`	✅	`imgsz`, `half`, `int8`, `nms`, `batch`, `data`
TF Edge TPU	`edgetpu`	`yolo11n_edgetpu.tflite`	✅	`imgsz`
TF.js	`tfjs`	`yolo11n_web_model/`	✅	`imgsz`, `half`, `int8`, `nms`, `batch`
PaddlePaddle	`paddle`	`yolo11n_paddle_model/`	✅	`imgsz`, `batch`
MNN	`mnn`	`yolo11n.mnn`	✅	`imgsz`, `batch`, `int8`, `half`
NCNN	`ncnn`	`yolo11n_ncnn_model/`	✅	`imgsz`, `half`, `batch`
IMX500	`imx`	`yolov8n_imx_model/`	✅	`imgsz`, `int8`, `data`
RKNN	`rknn`	`yolo11n_rknn_model/`	✅	`imgsz`, `batch`, `name`

Exciting New Features on the Way 🎉

We are continuously working to expand Ultralytics HUB's integration capabilities. Upcoming features include:

Additional Dataset Integrations
Detailed Export Integration Guides
Step-by-Step Tutorials for Each Integration

Stay Updated 🚧

Keep an eye on this integrations page for the latest developments. Stay informed through our:

Newsletter: Subscribe here for major announcements.
Social Media: Follow us on LinkedIn for updates and insights.
Blog: Visit our blog for detailed articles and use cases.

We Value Your Input 🗣️

Your feedback is crucial in shaping the future of Ultralytics HUB. Share your thoughts, suggestions, and integration requests here.

Thank You, Community! 🌍

Your contributions and engagement inspire our continuous innovation in AI and machine learning. Stay tuned for more exciting updates from Ultralytics!

📅 Created 1 year ago ✏️ Updated 1 day ago