flant5-dolly-QnA / README.md
perfectlybaked's picture
README updated
5ac691f
|
raw
history blame
450 Bytes
metadata
datasets:
  - databricks/databricks-dolly-15k
language:
  - en
metrics:
  - rouge
pipeline_tag: question-answering
tags:
  - question-answering

With the on-set of ChatGPT like products, there is a need of a question-answering model. Here we have finetuned Flan-T5 on a question answering dataset, where input is given as following: 'Context: <insert context for Q&A> \nInput: ' Model is trained on databricks dolly-15k dataset for QnA.