This release contains the Barista Datasets (Irfan et al. 2020; 2021) for generic and personalised task-oriented dialogue, in addition to the adapted scripts for running state-of-the-art data-driven dialogue models on the datasets: Supervised Embeddings (Dodge et al., 2015; Bordes et al., 2016), Sequence-to-Sequence (Sutskever et al., 2015), End-to-End Memory Networks (Sukhbaatar et al., 2016; Bordes et al., 2016), Generative Profile Memory Networks (Zhang et al., 2018), Key-Value Memory Networks (Miller et al., 2016; Zhang et al., 2018) and Split Memory Networks (Joshi et al., 2017).