Nous Research Releases NousCoder-14B: A Competitive Olympiad Programming Model Post-Trained on Qwen3-14B via Reinforcement Learning