model update
Browse files- README.md +111 -0
- eval/{metric.first.answer.paragraph_answer.question.asahi417_qg_ruquad.default.json → metric.first.answer.paragraph_answer.question.lmqg_qg_ruquad.default.json} +0 -0
- eval/{metric.first.answer.paragraph_sentence.question.asahi417_qg_ruquad.default.json → metric.first.answer.paragraph_sentence.question.lmqg_qg_ruquad.default.json} +0 -0
- eval/{metric.first.answer.sentence_answer.question.asahi417_qg_ruquad.default.json → metric.first.answer.sentence_answer.question.lmqg_qg_ruquad.default.json} +0 -0
- eval/{metric.first.sentence.paragraph_answer.question.asahi417_qg_ruquad.default.json → metric.first.sentence.paragraph_answer.question.lmqg_qg_ruquad.default.json} +0 -0
- eval/{metric.first.sentence.paragraph_sentence.question.asahi417_qg_ruquad.default.json → metric.first.sentence.paragraph_sentence.question.lmqg_qg_ruquad.default.json} +0 -0
- eval/{metric.first.sentence.sentence_answer.question.asahi417_qg_ruquad.default.json → metric.first.sentence.sentence_answer.question.lmqg_qg_ruquad.default.json} +0 -0
- eval/{metric.last.sentence.paragraph_answer.question.asahi417_qg_ruquad.default.json → metric.last.sentence.paragraph_answer.question.lmqg_qg_ruquad.default.json} +0 -0
- eval/{metric.last.sentence.paragraph_sentence.question.asahi417_qg_ruquad.default.json → metric.last.sentence.paragraph_sentence.question.lmqg_qg_ruquad.default.json} +0 -0
- eval/{metric.last.sentence.sentence_answer.question.asahi417_qg_ruquad.default.json → metric.last.sentence.sentence_answer.question.lmqg_qg_ruquad.default.json} +0 -0
- eval/{metric.long.sentence.paragraph_answer.question.asahi417_qg_ruquad.default.json → metric.long.sentence.paragraph_answer.question.lmqg_qg_ruquad.default.json} +0 -0
- eval/{metric.long.sentence.paragraph_sentence.question.asahi417_qg_ruquad.default.json → metric.long.sentence.paragraph_sentence.question.lmqg_qg_ruquad.default.json} +0 -0
- eval/{metric.long.sentence.sentence_answer.question.asahi417_qg_ruquad.default.json → metric.long.sentence.sentence_answer.question.lmqg_qg_ruquad.default.json} +0 -0
- eval/{metric.middle.sentence.paragraph_answer.question.asahi417_qg_ruquad.default.json → metric.middle.sentence.paragraph_answer.question.lmqg_qg_ruquad.default.json} +0 -0
- eval/{metric.middle.sentence.paragraph_sentence.question.asahi417_qg_ruquad.default.json → metric.middle.sentence.paragraph_sentence.question.lmqg_qg_ruquad.default.json} +0 -0
- eval/{metric.middle.sentence.sentence_answer.question.asahi417_qg_ruquad.default.json → metric.middle.sentence.sentence_answer.question.lmqg_qg_ruquad.default.json} +0 -0
- eval/{metric.short.sentence.paragraph_answer.question.asahi417_qg_ruquad.default.json → metric.short.sentence.paragraph_answer.question.lmqg_qg_ruquad.default.json} +0 -0
- eval/{metric.short.sentence.paragraph_sentence.question.asahi417_qg_ruquad.default.json → metric.short.sentence.paragraph_sentence.question.lmqg_qg_ruquad.default.json} +0 -0
- eval/{metric.short.sentence.sentence_answer.question.asahi417_qg_ruquad.default.json → metric.short.sentence.sentence_answer.question.lmqg_qg_ruquad.default.json} +0 -0
- eval/{samples.test.hyp.paragraph_answer.question.asahi417_qg_ruquad.default.txt → samples.test.hyp.paragraph_answer.question.lmqg_qg_ruquad.default.txt} +0 -0
- eval/{samples.test.hyp.paragraph_sentence.question.asahi417_qg_ruquad.default.txt → samples.test.hyp.paragraph_sentence.question.lmqg_qg_ruquad.default.txt} +0 -0
- eval/{samples.test.hyp.sentence_answer.question.asahi417_qg_ruquad.default.txt → samples.test.hyp.sentence_answer.question.lmqg_qg_ruquad.default.txt} +0 -0
- eval/{samples.validation.hyp.paragraph_answer.question.asahi417_qg_ruquad.default.txt → samples.validation.hyp.paragraph_answer.question.lmqg_qg_ruquad.default.txt} +0 -0
- eval/{samples.validation.hyp.paragraph_sentence.question.asahi417_qg_ruquad.default.txt → samples.validation.hyp.paragraph_sentence.question.lmqg_qg_ruquad.default.txt} +0 -0
- eval/{samples.validation.hyp.sentence_answer.question.asahi417_qg_ruquad.default.txt → samples.validation.hyp.sentence_answer.question.lmqg_qg_ruquad.default.txt} +0 -0
- trainer_config.json +1 -1
README.md
ADDED
@@ -0,0 +1,111 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
|
2 |
+
---
|
3 |
+
license: cc-by-4.0
|
4 |
+
metrics:
|
5 |
+
- bleu4
|
6 |
+
- meteor
|
7 |
+
- rouge-l
|
8 |
+
- bertscore
|
9 |
+
- moverscore
|
10 |
+
language: ru
|
11 |
+
datasets:
|
12 |
+
- lmqg/qg_ruquad
|
13 |
+
pipeline_tag: text2text-generation
|
14 |
+
tags:
|
15 |
+
- question generation
|
16 |
+
widget:
|
17 |
+
- text: "generate question: Нелишним будет отметить, что, развивая это направление, Д. И. Менделеев, поначалу априорно выдвинув идею о температуре, при которой высота мениска будет нулевой, <hl> в мае 1860 года <hl> провёл серию опытов."
|
18 |
+
example_title: "Question Generation Example 1"
|
19 |
+
- text: "generate question: Однако, франкоязычный <hl> Квебек <hl> практически никогда не включается в состав Латинской Америки."
|
20 |
+
example_title: "Question Generation Example 2"
|
21 |
+
- text: "generate question: Классическим примером международного синдиката XX века была группа компаний <hl> Де Бирс <hl> , которая в 1980-е годы контролировала до 90 % мировой торговли алмазами."
|
22 |
+
example_title: "Question Generation Example 3"
|
23 |
+
model-index:
|
24 |
+
- name: lmqg/mt5-small-ruquad
|
25 |
+
results:
|
26 |
+
- task:
|
27 |
+
name: Text2text Generation
|
28 |
+
type: text2text-generation
|
29 |
+
dataset:
|
30 |
+
name: lmqg/qg_ruquad
|
31 |
+
type: default
|
32 |
+
args: default
|
33 |
+
metrics:
|
34 |
+
- name: BLEU4
|
35 |
+
type: bleu4
|
36 |
+
value: 0.16312500044789852
|
37 |
+
- name: ROUGE-L
|
38 |
+
type: rouge-l
|
39 |
+
value: 0.31387559280955535
|
40 |
+
- name: METEOR
|
41 |
+
type: meteor
|
42 |
+
value: 0.2639298094863568
|
43 |
+
- name: BERTScore
|
44 |
+
type: bertscore
|
45 |
+
value: 0.8426647576680717
|
46 |
+
- name: MoverScore
|
47 |
+
type: moverscore
|
48 |
+
value: 0.6249344299293247
|
49 |
+
---
|
50 |
+
|
51 |
+
# Language Models Fine-tuning on Question Generation: `lmqg/mt5-small-ruquad`
|
52 |
+
This model is fine-tuned version of [google/mt5-small](https://huggingface.co/google/mt5-small) for question generation task on the
|
53 |
+
[lmqg/qg_ruquad](https://huggingface.co/datasets/lmqg/qg_ruquad) (dataset_name: default).
|
54 |
+
|
55 |
+
|
56 |
+
### Overview
|
57 |
+
- **Language model:** [google/mt5-small](https://huggingface.co/google/mt5-small)
|
58 |
+
- **Language:** ru
|
59 |
+
- **Training data:** [lmqg/qg_ruquad](https://huggingface.co/datasets/lmqg/qg_ruquad) (default)
|
60 |
+
- **Online Demo:** [https://autoqg.net/](https://autoqg.net/)
|
61 |
+
- **Repository:** [https://github.com/asahi417/lm-question-generation](https://github.com/asahi417/lm-question-generation)
|
62 |
+
- **Paper:** [TBA](TBA)
|
63 |
+
|
64 |
+
### Usage
|
65 |
+
```python
|
66 |
+
|
67 |
+
from transformers import pipeline
|
68 |
+
|
69 |
+
model_path = 'lmqg/mt5-small-ruquad'
|
70 |
+
pipe = pipeline("text2text-generation", model_path)
|
71 |
+
|
72 |
+
# Question Generation
|
73 |
+
input_text = 'generate question: Нелишним будет отметить, что, развивая это направление, Д. И. Менделеев, поначалу априорно выдвинув идею о температуре, при которой высота мениска будет нулевой, <hl> в мае 1860 года <hl> провёл серию опытов.'
|
74 |
+
question = pipe(input_text)
|
75 |
+
```
|
76 |
+
|
77 |
+
## Evaluation Metrics
|
78 |
+
|
79 |
+
|
80 |
+
### Metrics
|
81 |
+
|
82 |
+
| Dataset | Type | BLEU4 | ROUGE-L | METEOR | BERTScore | MoverScore | Link |
|
83 |
+
|:--------|:-----|------:|--------:|-------:|----------:|-----------:|-----:|
|
84 |
+
| [lmqg/qg_ruquad](https://huggingface.co/datasets/lmqg/qg_ruquad) | default | 0.16312500044789852 | 0.31387559280955535 | 0.2639298094863568 | 0.8426647576680717 | 0.6249344299293247 | [link](https://huggingface.co/lmqg/mt5-small-ruquad/raw/main/eval/metric.first.sentence.paragraph_answer.question.lmqg_qg_ruquad.default.json) |
|
85 |
+
|
86 |
+
|
87 |
+
|
88 |
+
|
89 |
+
## Training hyperparameters
|
90 |
+
|
91 |
+
The following hyperparameters were used during fine-tuning:
|
92 |
+
- dataset_path: lmqg/qg_ruquad
|
93 |
+
- dataset_name: default
|
94 |
+
- input_types: ['paragraph_answer']
|
95 |
+
- output_types: ['question']
|
96 |
+
- prefix_types: None
|
97 |
+
- model: google/mt5-small
|
98 |
+
- max_length: 512
|
99 |
+
- max_length_output: 32
|
100 |
+
- epoch: 5
|
101 |
+
- batch: 64
|
102 |
+
- lr: 0.001
|
103 |
+
- fp16: False
|
104 |
+
- random_seed: 1
|
105 |
+
- gradient_accumulation_steps: 1
|
106 |
+
- label_smoothing: 0.15
|
107 |
+
|
108 |
+
The full configuration can be found at [fine-tuning config file](https://huggingface.co/lmqg/mt5-small-ruquad/raw/main/trainer_config.json).
|
109 |
+
|
110 |
+
## Citation
|
111 |
+
TBA
|
eval/{metric.first.answer.paragraph_answer.question.asahi417_qg_ruquad.default.json → metric.first.answer.paragraph_answer.question.lmqg_qg_ruquad.default.json}
RENAMED
File without changes
|
eval/{metric.first.answer.paragraph_sentence.question.asahi417_qg_ruquad.default.json → metric.first.answer.paragraph_sentence.question.lmqg_qg_ruquad.default.json}
RENAMED
File without changes
|
eval/{metric.first.answer.sentence_answer.question.asahi417_qg_ruquad.default.json → metric.first.answer.sentence_answer.question.lmqg_qg_ruquad.default.json}
RENAMED
File without changes
|
eval/{metric.first.sentence.paragraph_answer.question.asahi417_qg_ruquad.default.json → metric.first.sentence.paragraph_answer.question.lmqg_qg_ruquad.default.json}
RENAMED
File without changes
|
eval/{metric.first.sentence.paragraph_sentence.question.asahi417_qg_ruquad.default.json → metric.first.sentence.paragraph_sentence.question.lmqg_qg_ruquad.default.json}
RENAMED
File without changes
|
eval/{metric.first.sentence.sentence_answer.question.asahi417_qg_ruquad.default.json → metric.first.sentence.sentence_answer.question.lmqg_qg_ruquad.default.json}
RENAMED
File without changes
|
eval/{metric.last.sentence.paragraph_answer.question.asahi417_qg_ruquad.default.json → metric.last.sentence.paragraph_answer.question.lmqg_qg_ruquad.default.json}
RENAMED
File without changes
|
eval/{metric.last.sentence.paragraph_sentence.question.asahi417_qg_ruquad.default.json → metric.last.sentence.paragraph_sentence.question.lmqg_qg_ruquad.default.json}
RENAMED
File without changes
|
eval/{metric.last.sentence.sentence_answer.question.asahi417_qg_ruquad.default.json → metric.last.sentence.sentence_answer.question.lmqg_qg_ruquad.default.json}
RENAMED
File without changes
|
eval/{metric.long.sentence.paragraph_answer.question.asahi417_qg_ruquad.default.json → metric.long.sentence.paragraph_answer.question.lmqg_qg_ruquad.default.json}
RENAMED
File without changes
|
eval/{metric.long.sentence.paragraph_sentence.question.asahi417_qg_ruquad.default.json → metric.long.sentence.paragraph_sentence.question.lmqg_qg_ruquad.default.json}
RENAMED
File without changes
|
eval/{metric.long.sentence.sentence_answer.question.asahi417_qg_ruquad.default.json → metric.long.sentence.sentence_answer.question.lmqg_qg_ruquad.default.json}
RENAMED
File without changes
|
eval/{metric.middle.sentence.paragraph_answer.question.asahi417_qg_ruquad.default.json → metric.middle.sentence.paragraph_answer.question.lmqg_qg_ruquad.default.json}
RENAMED
File without changes
|
eval/{metric.middle.sentence.paragraph_sentence.question.asahi417_qg_ruquad.default.json → metric.middle.sentence.paragraph_sentence.question.lmqg_qg_ruquad.default.json}
RENAMED
File without changes
|
eval/{metric.middle.sentence.sentence_answer.question.asahi417_qg_ruquad.default.json → metric.middle.sentence.sentence_answer.question.lmqg_qg_ruquad.default.json}
RENAMED
File without changes
|
eval/{metric.short.sentence.paragraph_answer.question.asahi417_qg_ruquad.default.json → metric.short.sentence.paragraph_answer.question.lmqg_qg_ruquad.default.json}
RENAMED
File without changes
|
eval/{metric.short.sentence.paragraph_sentence.question.asahi417_qg_ruquad.default.json → metric.short.sentence.paragraph_sentence.question.lmqg_qg_ruquad.default.json}
RENAMED
File without changes
|
eval/{metric.short.sentence.sentence_answer.question.asahi417_qg_ruquad.default.json → metric.short.sentence.sentence_answer.question.lmqg_qg_ruquad.default.json}
RENAMED
File without changes
|
eval/{samples.test.hyp.paragraph_answer.question.asahi417_qg_ruquad.default.txt → samples.test.hyp.paragraph_answer.question.lmqg_qg_ruquad.default.txt}
RENAMED
File without changes
|
eval/{samples.test.hyp.paragraph_sentence.question.asahi417_qg_ruquad.default.txt → samples.test.hyp.paragraph_sentence.question.lmqg_qg_ruquad.default.txt}
RENAMED
File without changes
|
eval/{samples.test.hyp.sentence_answer.question.asahi417_qg_ruquad.default.txt → samples.test.hyp.sentence_answer.question.lmqg_qg_ruquad.default.txt}
RENAMED
File without changes
|
eval/{samples.validation.hyp.paragraph_answer.question.asahi417_qg_ruquad.default.txt → samples.validation.hyp.paragraph_answer.question.lmqg_qg_ruquad.default.txt}
RENAMED
File without changes
|
eval/{samples.validation.hyp.paragraph_sentence.question.asahi417_qg_ruquad.default.txt → samples.validation.hyp.paragraph_sentence.question.lmqg_qg_ruquad.default.txt}
RENAMED
File without changes
|
eval/{samples.validation.hyp.sentence_answer.question.asahi417_qg_ruquad.default.txt → samples.validation.hyp.sentence_answer.question.lmqg_qg_ruquad.default.txt}
RENAMED
File without changes
|
trainer_config.json
CHANGED
@@ -1 +1 @@
|
|
1 |
-
{"dataset_path": "
|
|
|
1 |
+
{"dataset_path": "lmqg/qg_ruquad", "dataset_name": "default", "input_types": ["paragraph_answer"], "output_types": ["question"], "prefix_types": null, "model": "google/mt5-small", "max_length": 512, "max_length_output": 32, "epoch": 5, "batch": 64, "lr": 0.001, "fp16": false, "random_seed": 1, "gradient_accumulation_steps": 1, "label_smoothing": 0.15}
|