英文

lmqg/t5-small-squad-qag的模型卡片

这个模型是基于 t5-small 进行微调的,用于 lmqg 上的问答对生成任务(数据集名称:默认)。

概述

使用方式

from lmqg import TransformersQG

# initialize model
model = TransformersQG(language="en", model="lmqg/t5-small-squad-qag")

# model prediction
question_answer_pairs = model.generate_qa("William Turner was an English painter who specialised in watercolour landscapes")
  • 使用transformers
from transformers import pipeline

pipe = pipeline("text2text-generation", "lmqg/t5-small-squad-qag")
output = pipe("generate question and answer: Beyonce further expanded her acting career, starring as blues singer Etta James in the 2008 musical biopic, Cadillac Records.")

评估

Score Type Dataset
QAAlignedF1Score (BERTScore) 92.76 default 12313321
QAAlignedF1Score (MoverScore) 64.59 default 12313321
QAAlignedPrecision (BERTScore) 92.87 default 12313321
QAAlignedPrecision (MoverScore) 65.3 default 12313321
QAAlignedRecall (BERTScore) 92.68 default 12313321
QAAlignedRecall (MoverScore) 63.99 default 12313321

训练超参数

在微调过程中使用了以下超参数:

  • 数据集路径:lmqg/qag_squad
  • 数据集名称:默认
  • 输入类型:['paragraph']
  • 输出类型:['questions_answers']
  • 前缀类型:['qag']
  • 模型:t5-small
  • 最大长度:512
  • 输出最大长度:256
  • 迭代轮数:18
  • 批次大小:32
  • 学习率:0.0001
  • fp16:False
  • 随机种子:1
  • 梯度累积步数:2
  • 标签平滑:0.0

详细配置信息可在 fine-tuning config file 中找到。

引用

@inproceedings{ushio-etal-2022-generative,
    title = "{G}enerative {L}anguage {M}odels for {P}aragraph-{L}evel {Q}uestion {G}eneration",
    author = "Ushio, Asahi  and
        Alva-Manchego, Fernando  and
        Camacho-Collados, Jose",
    booktitle = "Proceedings of the 2022 Conference on Empirical Methods in Natural Language Processing",
    month = dec,
    year = "2022",
    address = "Abu Dhabi, U.A.E.",
    publisher = "Association for Computational Linguistics",
}