✨ Implement lazy loading for models and correct tokens counting

✨ Add Gemma 4 E2B model integration and update service to support multiple models

✨ Implement model gateway with Llama model integration and /generate API endpoint