Guides

Task-oriented guides for common workflows.

Guide	Description
Training	Configure and run model training
Training Guide	Configure and run model training
Config Generator	Generate training configs via TUI or auto mode
Negative Frames	Reduce false positives with background frames
Monitoring	WandB, visualizations, epoch-end evaluation
Multi-GPU	Scale training across multiple GPUs
Resume & Fine-Tune	Continue from existing weights
Supervised ID	Predict pose + persistent identity in one model
Inference	Run predictions on videos and label files
Running Inference	Run predictions from the CLI
Post-Processing Filters	Node-count, confidence, and overlap-NMS filters
Python API	Predictor / predict / Outputs for embedding inference in code
Centroid-Only	Run a centroid model standalone
Top-Down Segmentation	Per-instance masks via centroid + crop-mask
SAM-Prompted Segmentation	Per-instance masks from poses via Segment Anything
Performance	Tune inference throughput (FP16, torch.compile, workers)
Evaluation	Assess model performance with metrics
Tracking	Assign consistent IDs across frames
Export	ONNX/TensorRT for production inference

---

Looking for step-by-step learning? Check out the Tutorials.