We introduce an open-source Python package that provides a data-driven approach to generating atmospheric turbulence based on the deep rapid distortion (DRD) model in- troduced in Keith et al 2021. The approach involves an operator regression problem that determines the optimal candidate in a family of physics-informed covariance kernels that preserve fluid flow properties such as mass conservation as well as turbulence statistics up to second order. The package is based on the Mann model, which admits three parameters that may also determined directly from data by DRD models: the Kolmogorov constant multiplied by the two-thirds power law for the rate of viscous dissipation, a turbulence length scale, and a non-dimensional time-scale related to the eddy-lifetime function. Additionally, DRDMannTurb allows users to easily fit the Mann model to turbulence characteristics by also approximating the function related to the lifetime of the eddies through deep neural networks. Finally, the framework enables users to automatically generate three- dimensional turbulent wind fields in a scalable, memory-efficient manner. A variety of example scenarios involving various eddy lifetimes, data interpolation and filtering, and turbulence field generation are provided as part of the overall documentation. This talk will be complemented by a presentation in MS080 - Advances in Turbulence Modeling using Nonlocal Derivatives, Implicit LES and Deep Learning that focuses on the physical modelling and turbulence generation.