File size: 7,717 Bytes
8ec030d
 
 
 
 
88b4520
8ec030d
 
 
88b4520
8ec030d
88b4520
8ec030d
 
 
 
 
 
88b4520
 
8ec030d
 
b0b5a76
 
 
 
 
 
 
8ec030d
 
 
 
b0b5a76
8ec030d
d8fdd5a
8ec030d
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
b0b5a76
8ec030d
b0b5a76
8ec030d
 
b0b5a76
8ec030d
b0b5a76
8ec030d
 
b0b5a76
 
 
 
8ec030d
 
b0b5a76
8ec030d
 
b0b5a76
 
 
 
 
 
 
 
 
 
 
8ec030d
 
 
 
 
 
 
b0b5a76
 
8ec030d
b0b5a76
8ec030d
 
 
 
 
 
 
 
b0b5a76
8ec030d
b0b5a76
8ec030d
 
b0b5a76
 
 
 
8ec030d
 
b0b5a76
8ec030d
 
b0b5a76
 
 
 
 
 
 
 
 
 
 
8ec030d
 
 
 
 
 
 
b0b5a76
 
8ec030d
 
 
b0b5a76
 
8ec030d
 
 
 
 
b0b5a76
8ec030d
 
 
 
 
 
 
 
 
 
 
 
b0b5a76
8ec030d
 
 
 
 
 
 
 
 
b0b5a76
8ec030d
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
170
171
172
173
174
175
176
177
178
179
---
language:
- ar
- ca
- de
- el
- en
- es
- fr
- hi
- it
- ja
- ko
- nl
- pl
- pt
- ru
- sv
- vi
- zh
widget:
- text: >-
    I Red Hot Chili Peppers sono stati formati a Los Angeles da Kiedis, Flea, il chitarrista Hillel Slovak e il batterista Jack Irons.
  example_title: "Italian"
inference:
  parameters: 
    decoder_start_token_id: 250058
    src_lang: "it_XX"
    tgt_lang: "<triplet>"
tags:
- seq2seq
- relation-extraction
license: cc-by-nc-sa-4.0
pipeline_tag: translation
---
# RED<sup>FM</sup>: a Filtered and Multilingual Relation Extraction Dataset

This a multilingual version of [REBEL](https://huggingface.co/Babelscape/rebel-large). It can be used as a standalone multulingual Relation Extraction system, or as a pretrained system to be tuned on multilingual Relation Extraction datasets.

mREBEL is introduced in the ACL 2023 paper [RED^{FM}: a Filtered and Multilingual Relation Extraction Dataset](https://github.com/Babelscape/rebel/blob/main/docs/). We present a new multilingual Relation Extraction dataset and train a multilingual version of REBEL which reframed Relation Extraction as a seq2seq task. The paper can be found [here](https://github.com/Babelscape/rebel/blob/main/docs/). If you use the code or model, please reference this work in your paper:

    @inproceedings{huguet-cabot-et-al-2023-red,
        title = "RED^{FM}: a Filtered and Multilingual Relation Extraction Dataset",
        author = "Huguet Cabot, Pere-Llu{\'\i}s  and
          Navigli, Roberto",
        booktitle = "ACL 2023",
        month = jul,
        year = "2023",
        address = "Toronto, Canada",
        publisher = "Association for Computational Linguistics",
    }

The original repository for the paper can be found [here](https://github.com/Babelscape/rebel)

Be aware that the inference widget at the right does not output special tokens, which are necessary to distinguish the subject, object and relation types. For a demo of REBEL and its pre-training dataset check the [Spaces demo](https://huggingface.co/spaces/Babelscape/rebel-demo).

## Pipeline usage

```python
from transformers import pipeline

triplet_extractor = pipeline('translation_xx_to_yy', model='Babelscape/mrebel-large-32', tokenizer='Babelscape/mrebel-large-32')
# We need to use the tokenizer manually since we need special tokens.
extracted_text = triplet_extractor.tokenizer.batch_decode([triplet_extractor("The Red Hot Chili Peppers were formed in Los Angeles by Kiedis, Flea, guitarist Hillel Slovak and drummer Jack Irons.", decoder_start_token_id=250058, src_lang="en_XX", tgt_lang="<triplet>", return_tensors=True, return_text=False)[0]["translation_token_ids"]]) # change en_XX for the language of the source.
print(extracted_text[0])
# Function to parse the generated text and extract the triplets
def extract_triplets_typed(text):
    triplets = []
    relation = ''
    text = text.strip()
    current = 'x'
    subject, relation, object_, object_type, subject_type = '','','','',''

    for token in text.replace("<s>", "").replace("<pad>", "").replace("</s>", "").replace("tp_XX", "").replace("__en__", "").split():
        if token == "<triplet>" or token == "<relation>":
            current = 't'
            if relation != '':
                triplets.append({'head': subject.strip(), 'head_type': subject_type, 'type': relation.strip(),'tail': object_.strip(), 'tail_type': object_type})
                relation = ''
            subject = ''
        elif token.startswith("<") and token.endswith(">"):
            if current == 't' or current == 'o':
                current = 's'
                if relation != '':
                    triplets.append({'head': subject.strip(), 'head_type': subject_type, 'type': relation.strip(),'tail': object_.strip(), 'tail_type': object_type})
                object_ = ''
                subject_type = token[1:-1]
            else:
                current = 'o'
                object_type = token[1:-1]
                relation = ''
        else:
            if current == 't':
                subject += ' ' + token
            elif current == 's':
                object_ += ' ' + token
            elif current == 'o':
                relation += ' ' + token
    if subject != '' and relation != '' and object_ != '' and object_type != '' and subject_type != '':
        triplets.append({'head': subject.strip(), 'head_type': subject_type, 'type': relation.strip(),'tail': object_.strip(), 'tail_type': object_type})
    return triplets
extracted_triplets = extract_triplets_typed(extracted_text[0])
print(extracted_triplets)
```

## Model and Tokenizer using transformers

```python
from transformers import AutoModelForSeq2SeqLM, AutoTokenizer

def extract_triplets_typed(text):
    triplets = []
    relation = ''
    text = text.strip()
    current = 'x'
    subject, relation, object_, object_type, subject_type = '','','','',''

    for token in text.replace("<s>", "").replace("<pad>", "").replace("</s>", "").replace("tp_XX", "").replace("__en__", "").split():
        if token == "<triplet>" or token == "<relation>":
            current = 't'
            if relation != '':
                triplets.append({'head': subject.strip(), 'head_type': subject_type, 'type': relation.strip(),'tail': object_.strip(), 'tail_type': object_type})
                relation = ''
            subject = ''
        elif token.startswith("<") and token.endswith(">"):
            if current == 't' or current == 'o':
                current = 's'
                if relation != '':
                    triplets.append({'head': subject.strip(), 'head_type': subject_type, 'type': relation.strip(),'tail': object_.strip(), 'tail_type': object_type})
                object_ = ''
                subject_type = token[1:-1]
            else:
                current = 'o'
                object_type = token[1:-1]
                relation = ''
        else:
            if current == 't':
                subject += ' ' + token
            elif current == 's':
                object_ += ' ' + token
            elif current == 'o':
                relation += ' ' + token
    if subject != '' and relation != '' and object_ != '' and object_type != '' and subject_type != '':
        triplets.append({'head': subject.strip(), 'head_type': subject_type, 'type': relation.strip(),'tail': object_.strip(), 'tail_type': object_type})
    return triplets

# Load model and tokenizer
tokenizer = AutoTokenizer.from_pretrained("Babelscape/mrebel-large-32", src_lang="en_XX", "tgt_lang": "tp_XX") # Here we set English as source language. To change the source language just change it here or swap the first token of the input for your desired language
model = AutoModelForSeq2SeqLM.from_pretrained("Babelscape/mrebel-large-32")
gen_kwargs = {
    "max_length": 256,
    "length_penalty": 0,
    "num_beams": 3,
    "num_return_sequences": 3,
    "forced_bos_token_id": None,
}

# Text to extract triplets from
text = 'The Red Hot Chili Peppers were formed in Los Angeles by Kiedis, Flea, guitarist Hillel Slovak and drummer Jack Irons.'

# Tokenizer text
model_inputs = tokenizer(text, max_length=256, padding=True, truncation=True, return_tensors = 'pt')

# Generate
generated_tokens = model.generate(
    model_inputs["input_ids"].to(model.device),
    attention_mask=model_inputs["attention_mask"].to(model.device),
    decoder_start_token_id = self.tokenizer.convert_tokens_to_ids("tp_XX"),
    **gen_kwargs,
)

# Extract text
decoded_preds = tokenizer.batch_decode(generated_tokens, skip_special_tokens=False)

# Extract triplets
for idx, sentence in enumerate(decoded_preds):
    print(f'Prediction triplets sentence {idx}')
    print(extract_triplets_typed(sentence))
```