Even after running: copilot-proxy config set-context-length 200000 When I add Ollama to the copilot config, it shows all GLM models with 64k context, any idea on how to fix it?