Recent advancements in Additive Manufacturing have significantly increased the importance of multiscale topology optimization while facilitating its seamless integration. One of the methods for optimizing multiscale structure is concurrent approaches, in which the microscale topology optimization problem is solved at each Gaussian point of macrostructure in addition to macroscale topology optimization. However, this approach has high computational costs, making it impractical for large-scale structural applications. To address this problem, machine learning-based multiscale topology optimization has been studied. One of these methods uses predefined microstructures, such as lattice structures and triply periodic minimal surfaces, and restricts the design variables to parameters such as orientation, density, and geometric dimensions. A surrogate homogenization model is then constructed based on these parameters, enabling efficient macroscale optimization with reduced computational cost. However, to the best of our knowledge, most existing studies assume linear elasticity at both the micro and macro scales, with limited exploration of approaches incorporating finite deformation. In this study, we conduct a fundamental investigation toward multiscale topology optimization considering finite deformation. Specifically, we construct a surrogate model with a fixed microscopic geometry and apply it to macroscale topology optimization. The proposed method adopts the RBF-based surrogate homogenization model under finite deformation [1]. For optimization, we utilize the continuous adjoint method proposed by Han et al. [2] to perform the topology optimization of the macrostructure. As a representative numerical example, a stiffness maximization problem for a beam is conducted to demonstrate the appropriateness of the proposed method. [1] Nakamura A., Yamanaka Y., Nomura R., Moriguchi S., Terada K., Comput. Methods Appl. Mech. Engrg., Vol. 436, 2025 [2] Han J., Furuta K., Kondoh T., Nishiwaki S., Terada K., Comput. Methods Appl. Mech. Engrg., Vol. 429, 2024