add chat models: qwen/qwen2.5-coder-32b-instruct, qwen/qwen2.5-coder-7b-instruct, nvidia/llama-3.1-nemotron-70b-reward#128