1
Install Ultralytics Library
Run `pip install ultralytics` in a Python environment version 3.8 or higher with PyTorch 1.8 or above.
2
Upload Dataset
Upload images (up to 50MB), videos (up to 1GB), or ZIP files (up to 50GB) to the Ultralytics Platform.
3
Annotate Data
Use manual annotation tools or AI-assisted options like SAM smart annotation and YOLO auto-labeling.
4
Train Model on Cloud GPU
Select cloud GPU resources ranging from RTX 4090 to H200 and start training with real-time metrics monitoring.
5
Export and Deploy Model
Export trained models to one of 17 supported formats and deploy to one of 43 global regions using one-click endpoints.