File size: 7,868 Bytes
e6dc6f3
 
 
 
 
 
 
 
 
3197425
e6dc6f3
 
 
 
 
 
3197425
e6dc6f3
3197425
e6dc6f3
3197425
e6dc6f3
 
3197425
e6dc6f3
 
 
 
 
 
 
 
 
649297b
 
7fa6e03
649297b
 
7fa6e03
649297b
 
7fa6e03
649297b
 
7fa6e03
649297b
 
7fa6e03
98fadcb
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
e6dc6f3
 
3197425
7fa6e03
e6dc6f3
 
 
 
3197425
e6dc6f3
 
 
8aca83a
e6dc6f3
 
8aca83a
e6dc6f3
8aca83a
7fa6e03
8aca83a
3197425
7fa6e03
8aca83a
3197425
8aca83a
 
e6dc6f3
8aca83a
 
 
92eb337
3197425
 
e6dc6f3
7fa6e03
e6dc6f3
7fa6e03
e6dc6f3
 
3197425
e6dc6f3
7fa6e03
 
 
 
 
 
 
 
 
 
e6dc6f3
 
98fadcb
 
 
 
 
 
 
 
 
 
 
 
e6dc6f3
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
3197425
e6dc6f3
 
8aca83a
92eb337
8aca83a
92eb337
8aca83a
92eb337
 
 
 
 
 
 
 
8aca83a
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169

---
license: cc-by-4.0
metrics:
- bleu4
- meteor
- rouge-l
- bertscore
- moverscore
language: de
datasets:
- lmqg/qg_dequad
pipeline_tag: text2text-generation
tags:
- question generation
widget:
- text: "Empfangs- und Sendeantenne sollen in ihrer Polarisation übereinstimmen, andernfalls <hl> wird die Signalübertragung stark gedämpft. <hl>"
  example_title: "Question Generation Example 1" 
- text: "das erste weltweit errichtete Hermann Brehmer <hl> 1855 <hl> im niederschlesischen ''Görbersdorf'' (heute Sokołowsko, Polen)."
  example_title: "Question Generation Example 2" 
- text: "Er muss Zyperngrieche sein und wird direkt für <hl> fünf Jahre <hl> gewählt (Art. 43 Abs. 1 der Verfassung) und verfügt über weitreichende Exekutivkompetenzen."
  example_title: "Question Generation Example 3" 
model-index:
- name: lmqg/mt5-base-dequad-qg
  results:
  - task:
      name: Text2text Generation
      type: text2text-generation
    dataset:
      name: lmqg/qg_dequad
      type: default
      args: default
    metrics:
    - name: BLEU4 (Question Generation)
      type: bleu4_question_generation
      value: 0.87
    - name: ROUGE-L (Question Generation)
      type: rouge_l_question_generation
      value: 11.1
    - name: METEOR (Question Generation)
      type: meteor_question_generation
      value: 13.65
    - name: BERTScore (Question Generation)
      type: bertscore_question_generation
      value: 80.39
    - name: MoverScore (Question Generation)
      type: moverscore_question_generation
      value: 55.73
    - name: QAAlignedF1Score-BERTScore (Question & Answer Generation) [Gold Answer]
      type: qa_aligned_f1_score_bertscore_question_answer_generation_gold_answer
      value: 90.63
    - name: QAAlignedRecall-BERTScore (Question & Answer Generation) [Gold Answer]
      type: qa_aligned_recall_bertscore_question_answer_generation_gold_answer
      value: 90.61
    - name: QAAlignedPrecision-BERTScore (Question & Answer Generation) [Gold Answer]
      type: qa_aligned_precision_bertscore_question_answer_generation_gold_answer
      value: 90.65
    - name: QAAlignedF1Score-MoverScore (Question & Answer Generation) [Gold Answer]
      type: qa_aligned_f1_score_moverscore_question_answer_generation_gold_answer
      value: 65.32
    - name: QAAlignedRecall-MoverScore (Question & Answer Generation) [Gold Answer]
      type: qa_aligned_recall_moverscore_question_answer_generation_gold_answer
      value: 65.3
    - name: QAAlignedPrecision-MoverScore (Question & Answer Generation) [Gold Answer]
      type: qa_aligned_precision_moverscore_question_answer_generation_gold_answer
      value: 65.34
---

# Model Card of `lmqg/mt5-base-dequad-qg`
This model is fine-tuned version of [google/mt5-base](https://huggingface.co/google/mt5-base) for question generation task on the [lmqg/qg_dequad](https://huggingface.co/datasets/lmqg/qg_dequad) (dataset_name: default) via [`lmqg`](https://github.com/asahi417/lm-question-generation).


### Overview
- **Language model:** [google/mt5-base](https://huggingface.co/google/mt5-base)   
- **Language:** de  
- **Training data:** [lmqg/qg_dequad](https://huggingface.co/datasets/lmqg/qg_dequad) (default)
- **Online Demo:** [https://autoqg.net/](https://autoqg.net/)
- **Repository:** [https://github.com/asahi417/lm-question-generation](https://github.com/asahi417/lm-question-generation)
- **Paper:** [https://arxiv.org/abs/2210.03992](https://arxiv.org/abs/2210.03992)

### Usage
- With [`lmqg`](https://github.com/asahi417/lm-question-generation#lmqg-language-model-for-question-generation-)
```python
from lmqg import TransformersQG

# initialize model
model = TransformersQG(language="de", model="lmqg/mt5-base-dequad-qg")

# model prediction
questions = model.generate_q(list_context="das erste weltweit errichtete Hermann Brehmer 1855 im niederschlesischen ''Görbersdorf'' (heute Sokołowsko, Polen).", list_answer="1855")

```

- With `transformers`
```python
from transformers import pipeline

pipe = pipeline("text2text-generation", "lmqg/mt5-base-dequad-qg")
output = pipe("Empfangs- und Sendeantenne sollen in ihrer Polarisation übereinstimmen, andernfalls <hl> wird die Signalübertragung stark gedämpft. <hl>")

```

## Evaluation


- ***Metric (Question Generation)***: [raw metric file](https://huggingface.co/lmqg/mt5-base-dequad-qg/raw/main/eval/metric.first.sentence.paragraph_answer.question.lmqg_qg_dequad.default.json) 

|            |   Score | Type    | Dataset                                                          |
|:-----------|--------:|:--------|:-----------------------------------------------------------------|
| BERTScore  |   80.39 | default | [lmqg/qg_dequad](https://huggingface.co/datasets/lmqg/qg_dequad) |
| Bleu_1     |   10.85 | default | [lmqg/qg_dequad](https://huggingface.co/datasets/lmqg/qg_dequad) |
| Bleu_2     |    4.61 | default | [lmqg/qg_dequad](https://huggingface.co/datasets/lmqg/qg_dequad) |
| Bleu_3     |    2.06 | default | [lmqg/qg_dequad](https://huggingface.co/datasets/lmqg/qg_dequad) |
| Bleu_4     |    0.87 | default | [lmqg/qg_dequad](https://huggingface.co/datasets/lmqg/qg_dequad) |
| METEOR     |   13.65 | default | [lmqg/qg_dequad](https://huggingface.co/datasets/lmqg/qg_dequad) |
| MoverScore |   55.73 | default | [lmqg/qg_dequad](https://huggingface.co/datasets/lmqg/qg_dequad) |
| ROUGE_L    |   11.1  | default | [lmqg/qg_dequad](https://huggingface.co/datasets/lmqg/qg_dequad) |


- ***Metric (Question & Answer Generation)***: QAG metrics are computed with *the gold answer* and generated question on it for this model, as the model cannot provide an answer. [raw metric file](https://huggingface.co/lmqg/mt5-base-dequad-qg/raw/main/eval/metric.first.answer.paragraph.questions_answers.lmqg_qg_dequad.default.json)

|                                 |   Score | Type    | Dataset                                                          |
|:--------------------------------|--------:|:--------|:-----------------------------------------------------------------|
| QAAlignedF1Score (BERTScore)    |   90.63 | default | [lmqg/qg_dequad](https://huggingface.co/datasets/lmqg/qg_dequad) |
| QAAlignedF1Score (MoverScore)   |   65.32 | default | [lmqg/qg_dequad](https://huggingface.co/datasets/lmqg/qg_dequad) |
| QAAlignedPrecision (BERTScore)  |   90.65 | default | [lmqg/qg_dequad](https://huggingface.co/datasets/lmqg/qg_dequad) |
| QAAlignedPrecision (MoverScore) |   65.34 | default | [lmqg/qg_dequad](https://huggingface.co/datasets/lmqg/qg_dequad) |
| QAAlignedRecall (BERTScore)     |   90.61 | default | [lmqg/qg_dequad](https://huggingface.co/datasets/lmqg/qg_dequad) |
| QAAlignedRecall (MoverScore)    |   65.3  | default | [lmqg/qg_dequad](https://huggingface.co/datasets/lmqg/qg_dequad) |



## Training hyperparameters

The following hyperparameters were used during fine-tuning:
 - dataset_path: lmqg/qg_dequad
 - dataset_name: default
 - input_types: ['paragraph_answer']
 - output_types: ['question']
 - prefix_types: None
 - model: google/mt5-base
 - max_length: 512
 - max_length_output: 32
 - epoch: 17
 - batch: 4
 - lr: 0.0005
 - fp16: False
 - random_seed: 1
 - gradient_accumulation_steps: 16
 - label_smoothing: 0.15

The full configuration can be found at [fine-tuning config file](https://huggingface.co/lmqg/mt5-base-dequad-qg/raw/main/trainer_config.json).

## 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",
}

```