Train and export Ultralytics YOLO models for detection and tracking.
npx skills add https://github.com/jayll1303/AIEKit --skill ultralytics-yolo
Please help me install this Agent Skill. Skill: ultralytics-yolo Source: https://github.com/jayll1303/AIEKit/tree/main/.kiro/skills/ultralytics-yolo Command: npx skills add https://github.com/jayll1303/AIEKit --skill ultralytics-yolo
Simplifies the end-to-end workflow of training, validating, running inference, exporting, and deploying Ultralytics YOLO models so practitioners can move from raw images to production-ready models without guesswork.
Train a yolo26n model on my dataset using the data.yaml at path /path/to/data.yaml for 100 epochs and export the best checkpoint to ONNX.
High-intent search queries and answers about installing and using this skill.
Train YOLO models on custom datasets using a standardized data.yaml file, then export the best checkpoint to ONNX. You can specify training epochs, model size like yolo26n, and target devices like multi-GPU or Apple MPS for hardware acceleration.
Run YOLO inference on images, videos, and streams while performing multi-object tracking. It supports BoT-SORT and ByteTrack algorithms, providing programmatic access to results like boxes, masks, and keypoints for downstream analytics.
YOLO model export supports formats including TensorRT, TFLite, CoreML, and OpenVINO. You can apply quantization flags like FP16 or INT8 during export to optimize models for edge and server deployment environments.
YOLO object detection supports segmentation, classification, pose estimation, and oriented bounding boxes. You can train and validate models for these tasks, monitoring metrics like mAP, precision, and recall across CPU, GPU, or MPS devices.
Extract bounding boxes, masks, keypoints, and oriented bounding boxes from YOLO predictions programmatically. These results enable downstream processing and analytics solutions like object counting, tracking, and heatmaps.
A data.yaml file standardizes custom dataset preparation for YOLO model training. It defines class names and dataset paths, ensuring balanced training across CPU, GPU, and MPS devices while monitoring validation metrics like mAP.
KarlinskySClassifies images and computes image-text cosine similarity without training.
eddieluengGenerate object masks from point, box, or mask prompts via SAM.
SJTU-IPADSDesign, train, optimize, and deploy computer vision systems with detection, segmentation, and transformer models.
Fantasia1999Trains computer vision models for detection and segmentation with PyTorch and YOLO/CNN architectures.
