mit einem Klick
ml-dev-principles
General working discipline for ML / multimodal / CV development (how to work, not which library). Load at the START of an ML task โ model/dataset candidate selection, license or eligibility filtering, EDA, training/fine-tuning (SFT/DPO/LoRA), evaluation-harness and train/val/test isolation design, error analysis on FP/FN cases, or GPU throughput tuning. Triggers on "EDA", "explore dataset", "error analysis", "FP/FN", "train", "fine-tune", "SFT", "DPO", "LoRA", "eval harness", "validation split", "test leakage", "dataset selection", "model selection", "license filtering", "GPU utilization", "multimodal", "VLM", "image review".
Mit Codex oder Claude installieren Kopieren Sie diesen Prompt, fรผgen Sie ihn in Codex, Claude oder einen anderen Assistant ein und lassen Sie die Skill-Seite prรผfen und installieren.