Microsoft Research |

BioGPT: Question and Answer - PubMed

639
11
Bio LLMs
    More
PubMedQA (Standard)
Response
The response to your query will appear here

BioGPT: Question and Answer - PubMed is a language model fine-tuned on the expert annotated PubMedQA dataset which is a biomedical question-answering dataset, to answer yes/no/maybe questions for biomedical tasks. BioGPT outperforms larger, more general language models in biomedical language benchmarks.

Example use case:  Answering academic questions. See examples below.

Technology:  GPT-2 Backbone

Metrics: 78.2% Accuracy
  
Note: For better results, queries are separated into a question and a context for the question. See examples below.

Citation:
Renqian Luo, Liai Sun, Yingce Xia, Tao Qin, Sheng Zhang, Hoifung Poon, Tie-Yan Liu, BioGPT: generative pre-trained transformer for biomedical text generation and mining, Briefings in Bioinformatics, Volume 23, Issue 6, November 2022, bbac409, https://doi.org/10.1093/bib/bbac409
Released:
Feb-23-2023