nasa-smd-ibm-v0.1 is a RoBERTa-based encoder-decoder model fine-tuned for NASA science missions. It has been trained on scientific journals and articles related to NASA science tasks to enhance natural language processing capabilities such as information retrieval and intelligent search. This 125-million parameter model uses a masked language model for pre-training. It can be used for tasks like named entity recognition, information retrieval, sentence conversion, and extensible question answering, specifically targeting use cases related to NASA science missions.