lmqg/flan-t5-base-squad-qg 模型卡片

该模型是 google/flan-t5-base 的fine-tuned版本，用于 lmqg 数据集上的问题生成任务 (数据集名称：default)。

概述

• 语言模型： google/flan-t5-base
• 语言：en
• 训练数据： lmqg/qg_squad (默认)
• 在线演示： https://autoqg.net/
• 代码库： https://github.com/asahi417/lm-question-generation
• 论文： https://arxiv.org/abs/2210.03992

用法

• 使用 lmqg

from lmqg import TransformersQG

# initialize model
model = TransformersQG(language="en", model="lmqg/flan-t5-base-squad-qg")

# model prediction
questions = model.generate_q(list_context="William Turner was an English painter who specialised in watercolour landscapes", list_answer="William Turner")

• 使用 transformers

from transformers import pipeline

pipe = pipeline("text2text-generation", "lmqg/flan-t5-base-squad-qg")
output = pipe("generate question: <hl> Beyonce <hl> further expanded her acting career, starring as blues singer Etta James in the 2008 musical biopic, Cadillac Records.")

评估

• 评估指标 (问题生成)： raw metric file

• 评估指标 (问题和答案生成，管道方法)：每个问题是根据 lmqg/flan-t5-base-squad-ae 生成的答案生成的， raw metric file

训练超参数

以下是fine-tuning期间使用的超参数：

• 数据集路径：lmqg/qg_squad
• 数据集名称：default
• 输入类型：paragraph_answer
• 输出类型：question
• 前缀类型：['qg']
• 模型：google/flan-t5-base
• 最大长度：512
• 输出最大长度：32
• 迭代轮次：7
• batch大小：16
• 学习率：5e-05
• fp16：False
• 随机种子：1
• 渐变累积步数：4
• 标签平滑：0.15

完整的配置信息可以在 fine-tuning config file 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",
}