object-detection - YOLO Object Detection Fine-tuning
Fine-tune YOLO object detection models on custom datasets and export to ONNX
Tags
Updated: 2026-03-26Capabilities
Typical Inputs
Typical Outputs
What this skill does
- inspect dataset format
- convert dataset format
- prepare dataset splits
- install dependencies
- train YOLO model
- evaluate model metrics
- export ONNX model
- collect training artifacts
- generate training summary
Inputs
- custom dataset directory
- dataset configuration file
Outputs
- trained model weights
- evaluation report
- ONNX model file
- TensorBoard logs
- visualization charts
Requirements
- Bash
- Python >= 3.8
- ultralytics
- onnx
- opencv-python-headless