Computational models play a key role in elucidating the complex behaviour of neurons. However, many of these models are tuned to the electrophysiological features of adult neurons, leaving a significant gap in the modelling of immature perinatal neurons. This study aims to bridge this gap by generating a population of computational models representing hippocampal neurons from primary hippocampal cultures of neonatal rat brains and comparing them with models of adult neurons in terms of variability in ion channels and electrophysiological behaviour. Using the TREES toolbox, we reconstructed the morphology of neurons from stacks of fluorescence images. This allowed accurate representation of their dendritic structures. We used the NEURON software to simulate evoked neuronal activity, implementing hyperpolarising and depolarising current pulses, that mimics patch clamp experiments in vitro. We used deep evolutionary algorithms in Python (DAEP) to identify best parameter combinations that yielded plausible model fits. This work resulted in a set of in silico models of hippocampal neurons that reflect the variability observed in the experimental dataset. In the future, we aim to use this population as a baseline population for modelling data from neurons whose mothers were exposed to chronic unpredictable stress before gestation (maternal depression) and studying observed differences in excitability. Supported by a grant APVV-19-0435 and VEGA 2/0081/22.