Learning Foundation Language Models for Geoscience Knowledge Understanding and Utilization

Show simple item record

dc.contributor.author Deng C.
dc.contributor.author Zhang T.
dc.contributor.author He Z.
dc.contributor.author Chen Q.
dc.contributor.author Shi Y.
dc.contributor.author Zhou L.
dc.contributor.author Fu L.
dc.contributor.author Zhang W.
dc.contributor.author Wang X.
dc.contributor.author Zhou C.
dc.contributor.author Lin Z.
dc.contributor.author He J.
dc.date.accessioned 2023-07-26T12:13:36Z
dc.date.available 2023-07-26T12:13:36Z
dc.date.issued 2023
dc.identifier.citation arXiv:2306.05064, 2023 ru_RU
dc.identifier.uri https://repository.geologyscience.ru/handle/123456789/41601
dc.description.abstract Large language models (LLMs) have achieved great success in general domains of natural language processing. In this paper, we bring LLMs to the realm of geoscience with the objective of advancing research and applications in this field. To this end, we present the first-ever LLM in geoscience, K2, alongside a suite of resources developed to further promote LLM research within geoscience. For instance, we have curated the first geoscience instruction tuning dataset, GeoSignal, which aims to align LLM responses to geoscience-related user queries. Additionally, we have established the first geoscience benchmark, GeoBenchmark, to evaluate LLMs in the context of geoscience. In this work, we experiment with a complete recipe to adapt a pretrained general-domain LLM to the geoscience domain. Specifically, we further train the LLaMA-7B model on over 2 million pieces of geoscience literature (3.9B Tokens) and utilize GeoSignal’s supervised data to fine-tune the model. Moreover, we share a protocol that can efficiently gather domain-specific data and construct domain-supervised data, even in situations where manpower is scarce. Experiments conducted on the GeoBenchmark demonstrate the effectiveness of our approach and datasets. ru_RU
dc.language.iso en ru_RU
dc.subject Geoscience Language Model ru_RU
dc.subject Domain Adaptation ru_RU
dc.title Learning Foundation Language Models for Geoscience Knowledge Understanding and Utilization ru_RU
dc.type Article ru_RU


Files in this item

This item appears in the following Collection(s)

Show simple item record