YOLO-Master Agent Skill

Use This Skill

Use this skill for any repository task that should drive the YOLO-Master stack end-to-end:

• train, val, predict, track, export, benchmark, tune
• model inspection and task detection
• LoRA save/load/merge and yolo.lora.diagnose
• PEFT comparison via yolo.eval.peft_compare
• Sparse SAHI comparison via yolo.eval.sparse_sahi_compare
• MoE diagnose/prune
• multimodal visual inference with OpenAI VLM/LLM cooperation
• multimodal batch evaluation over a dataset or image folder
• solutions workflows
• launchers for Gradio / Streamlit
• end-to-end orchestration via yolo.pipeline.experiment

Execution Rule