This model is a Llama-2-7b-chat-hf model trained on ErAConD data, which is a dataset for grammar corrections for written dialogue conversations. This model is presented in Using Adaptive Empathetic Responses for Teaching English. Please refer to the paper and the corresponding repository for training detaills.