Modeling of disruptions and runaway electrons is crucial for the safe operation of fusion reactors. However, creating conditions representative of disruptions in reactor-scale devices in current experiments is not possible, calling for the development of reliable and validated simulation tools. The validation requires the knowledge of plasma parameters during disruptions, but they are poorly constrained due to the rapid timescales.\ In this work, we simulate the total plasma current evolution to better constrain some of the parameters, such as the assimilated fraction of material during massive gas injection, the heat diffusivity and the hyperdiffusivity[^1] of current density, that is responsible for the current density flattening and the associated current spike[^2]. To find the optimal values of the free parameters, Bayesian optimization was employed, which has proven to be efficient for multi-parameter optimization, based on the numerical framework developed by Järvinen et al.[^3]\ Our results show that the time scale of the injection and transport of argon cannot be neglected compared to the time scales of current and temperature evolution. We were able to successfully constrain the amount of assimilated argon, the hyperdiffusivity, and the time scale of argon assimilation, but obtained surprisingly small values for the magnetic perturbation amplitude characterizing the temperature evolution, with double minima. [^1]: I. Pusztai, M. Hoppe, and O. Vallhagen, “Runaway dynamics in tokamak disruptions with current relaxation,” Journal of Plasma Physics, vol. 88, no. 4, p. 905 880 409, 2022. [^2]: A. H. Boozer, “Pivotal issues on relativistic electrons in ITER,” Nuclear Fusion, vol. 58, no. 3, p. 036 006, Jan. 2018. [^3]: A. Järvinen, T. Fülöp, E. Hirvijoki, M. Hoppe, A. Kit, and J. Åström, “Bayesian approach for validation of runaway electron simulations,” Journal of Plasma Physics, vol. 88, no. 6, p. 905 880 612, 2022.