The simulative characterization of the creep response of short-fiber reinforced thermoplastics (SFRTs) is complex due to the presence of multiple scales, both in space (due to the reinforcements) and in time (short- and long-term effects). To reduce the involved effort, computational multiscale approaches may be used to calibrate these models experimentally. Since the various parameters of the model must be identified from experimental data, the calibration of the microscopic material models is time consuming. In the current work, we replace the micromechanical model with the deep material network (DMN) framework, which is a data-driven homogenization technique. The DMNs can reproduce the full-field simulations while using highly non-linear creep models and offers a speed-up factor of 600 on average [1]. Thereafter, we perform an inverse parameter optimization of the matrix model of a SFRT using DMNs [2]. Due to the substantial anisotropy and the extensive time scales involved, we are particularly interested in the long-term creep response of SFRTs, which presents unique difficulties for both experimental and simulation-based techniques. Furthermore, we are interested in a more generalized framework, wherein we use multiple trained DMNs to interpolate over different fiber orientation tensors and fiber lengths within the microstructure. The interpolation over fiber orientation and fiber length makes the simulation framework suitable not only for the calibration of various material models for different classes of SFRTs but also for component simulation. REFERENCES [1] Dey, A. P., Welschinger, F., Schneider, M., Gajek, S. & Böhlke, T. Training deep material networks to reproduce creep loading of short fiber-reinforced thermoplastics with an inelastically-informed strategy, Arch. Appl. Mech. 2022; 92, 2733–2755. [2] Dey, A. P., Welschinger, F., Schneider, M., Gajek, S. & Böhlke, Rapid inverse calibration of a multiscale model for the viscoplastic and creep behavior of short fiber-reinforced thermoplastics based on Deep Material Networks, Int. J. Plast 2022; 103484.