I will debug your Python AI model, YOLO, PyTorch or TensorFlow

Is your Python ML model failing to converge, returning the same prediction for every input, or producing unexpected results? I will diagnose one reproducible issue in your PyTorch, TensorFlow, YOLO, or OpenCV project. I can inspect data preprocessing, class labels, model output decoding, dependency versions, training logs and evaluation code. Starter package: review one codebase and issue; deliver a tested fix where feasible or root cause report; provide patch, run instructions and one scope-limited follow-up. Please provide sample inputs, logs, source code, and library versions. Delivery: three business days after agreed scope. GPU costs, large retraining, full product redesign, paid APIs and production deployment are not included. Message before ordering. Work begins once scope and escrow funding are confirmed.