metadata

license: cc-by-4.0
metrics:
  - bleu4
  - meteor
  - rouge-l
  - bertscore
  - moverscore
language: es
datasets:
  - lmqg/qag_esquad
pipeline_tag: text2text-generation
tags:
  - questions and answers generation
widget:
  - text: del Ministerio de Desarrollo Urbano , Gobierno de la India.
    example_title: Questions & Answers Generation Example 1
model-index:
  - name: lmqg/mt5-base-esquad-qag
    results:
      - task:
          name: Text2text Generation
          type: text2text-generation
        dataset:
          name: lmqg/qag_esquad
          type: default
          args: default
        metrics:
          - name: QAAlignedF1Score-BERTScore (Question & Answer Generation)
            type: qa_aligned_f1_score_bertscore_question_answer_generation
            value: 78.96
          - name: QAAlignedRecall-BERTScore (Question & Answer Generation)
            type: qa_aligned_recall_bertscore_question_answer_generation
            value: 79.31
          - name: QAAlignedPrecision-BERTScore (Question & Answer Generation)
            type: qa_aligned_precision_bertscore_question_answer_generation
            value: 78.66
          - name: QAAlignedF1Score-MoverScore (Question & Answer Generation)
            type: qa_aligned_f1_score_moverscore_question_answer_generation
            value: 54.3
          - name: QAAlignedRecall-MoverScore (Question & Answer Generation)
            type: qa_aligned_recall_moverscore_question_answer_generation
            value: 54.42
          - name: QAAlignedPrecision-MoverScore (Question & Answer Generation)
            type: qa_aligned_precision_moverscore_question_answer_generation
            value: 54.21

Model Card of `lmqg/mt5-base-esquad-qag`

This model is fine-tuned version of google/mt5-base for question & answer pair generation task on the lmqg/qag_esquad (dataset_name: default) via lmqg.

Overview

Language model: google/mt5-base
Language: es
Training data: lmqg/qag_esquad (default)
Online Demo: https://autoqg.net/
Repository: https://github.com/asahi417/lm-question-generation
Paper: https://arxiv.org/abs/2210.03992

Usage

With lmqg

from lmqg import TransformersQG

# initialize model
model = TransformersQG(language="es", model="lmqg/mt5-base-esquad-qag")

# model prediction
question_answer_pairs = model.generate_qa("a noviembre , que es también la estación lluviosa.")

With transformers

from transformers import pipeline

pipe = pipeline("text2text-generation", "lmqg/mt5-base-esquad-qag")
output = pipe("del Ministerio de Desarrollo Urbano , Gobierno de la India.")

Evaluation

Metric (Question & Answer Generation): raw metric file

	Score	Type	Dataset
QAAlignedF1Score (BERTScore)	78.96	default	lmqg/qag_esquad
QAAlignedF1Score (MoverScore)	54.3	default	lmqg/qag_esquad
QAAlignedPrecision (BERTScore)	78.66	default	lmqg/qag_esquad
QAAlignedPrecision (MoverScore)	54.21	default	lmqg/qag_esquad
QAAlignedRecall (BERTScore)	79.31	default	lmqg/qag_esquad
QAAlignedRecall (MoverScore)	54.42	default	lmqg/qag_esquad

Training hyperparameters

The following hyperparameters were used during fine-tuning:

dataset_path: lmqg/qag_esquad
dataset_name: default
input_types: ['paragraph']
output_types: ['questions_answers']
prefix_types: None
model: google/mt5-base
max_length: 512
max_length_output: 256
epoch: 13
batch: 2
lr: 0.0005
fp16: False
random_seed: 1
gradient_accumulation_steps: 32
label_smoothing: 0.0

The full configuration can be found at fine-tuning config file.

Citation

@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",
}