GALACTICA is a general-purpose scientific language model. It is trained on a large corpus of scientific text and data. It can perform scientific NLP tasks at a high level, as well as tasks such as citation prediction, mathematical reasoning, molecular property prediction and protein annotation. A demo is available at galactica.org.
GitHub – paperswithcode/galai: Model API for GALACTICA
par
Étiquettes :