Support code execution flag for Google Gemini 1.5 models

It's a simple flag and allows Gemini to generate and run code, reporting the output. I'd love to see this supported in TypingMind. https://ai.google.dev/gemini-api/docs/code-execution?lang=python