1 ---
2 language: en
3 pipeline_tag: fill-mask
4 license: cc-by-sa-4.0
5 thumbnail: https://i.ibb.co/p3kQ7Rw/Screenshot-2020-10-06-at-12-16-36-PM.png
6 tags:
7 - legal
8 widget:
9 - text: "The applicant submitted that her husband was subjected to treatment amounting to [MASK] whilst in the custody of police."
10 ---
11
12 # LEGAL-BERT: The Muppets straight out of Law School
13
14 <img align="left" src="https://i.ibb.co/p3kQ7Rw/Screenshot-2020-10-06-at-12-16-36-PM.png" width="100"/>
15
16 LEGAL-BERT is a family of BERT models for the legal domain, intended to assist legal NLP research, computational law, and legal technology applications. To pre-train the different variations of LEGAL-BERT, we collected 12 GB of diverse English legal text from several fields (e.g., legislation, court cases, contracts) scraped from publicly available resources. Sub-domains variants (CONTRACTS-, EURLEX-, ECHR-) and/or general LEGAL-BERT perform better than using BERT out of the box for domain-specific tasks. A light-weight model (33% the size of BERT-BASE) pre-trained from scratch on legal data with competitive perfomance is also available.
17 <br/><br/><br/><br/>
18
19 ---
20
21 I. Chalkidis, M. Fergadiotis, P. Malakasiotis, N. Aletras and I. Androutsopoulos. "LEGAL-BERT: The Muppets straight out of Law School". In Findings of Empirical Methods in Natural Language Processing (EMNLP 2020) (Short Papers), to be held online, 2020. (https://aclanthology.org/2020.findings-emnlp.261)
22
23 ---
24
25 ## Pre-training corpora
26
27 The pre-training corpora of LEGAL-BERT include:
28
29 * 116,062 documents of EU legislation, publicly available from EURLEX (http://eur-lex.europa.eu), the repository of EU Law running under the EU Publication Office.
30
31 * 61,826 documents of UK legislation, publicly available from the UK legislation portal (http://www.legislation.gov.uk).
32
33 * 19,867 cases from the European Court of Justice (ECJ), also available from EURLEX.
34
35 * 12,554 cases from HUDOC, the repository of the European Court of Human Rights (ECHR) (http://hudoc.echr.coe.int/eng).
36
37 * 164,141 cases from various courts across the USA, hosted in the Case Law Access Project portal (https://case.law).
38
39 * 76,366 US contracts from EDGAR, the database of US Securities and Exchange Commission (SECOM) (https://www.sec.gov/edgar.shtml).
40
41 ## Pre-training details
42
43 * We trained BERT using the official code provided in Google BERT's github repository (https://github.com/google-research/bert).
44 * We released a model similar to the English BERT-BASE model (12-layer, 768-hidden, 12-heads, 110M parameters).
45 * We chose to follow the same training set-up: 1 million training steps with batches of 256 sequences of length 512 with an initial learning rate 1e-4.
46 * We were able to use a single Google Cloud TPU v3-8 provided for free from [TensorFlow Research Cloud (TFRC)](https://www.tensorflow.org/tfrc), while also utilizing [GCP research credits](https://edu.google.com/programs/credits/research). Huge thanks to both Google programs for supporting us!
47 * Part of LEGAL-BERT is a light-weight model pre-trained from scratch on legal data, which achieves comparable performance to larger models, while being much more efficient (approximately 4 times faster) with a smaller environmental footprint.
48
49 ## Models list
50
51 | Model name | Model Path | Training corpora |
52 | ------------------- | ------------------------------------ | ------------------- |
53 | CONTRACTS-BERT-BASE | `nlpaueb/bert-base-uncased-contracts` | US contracts |
54 | EURLEX-BERT-BASE | `nlpaueb/bert-base-uncased-eurlex` | EU legislation |
55 | ECHR-BERT-BASE | `nlpaueb/bert-base-uncased-echr` | ECHR cases |
56 | LEGAL-BERT-BASE * | `nlpaueb/legal-bert-base-uncased` | All |
57 | LEGAL-BERT-SMALL | `nlpaueb/legal-bert-small-uncased` | All |
58
59 \* LEGAL-BERT-BASE is the model referred to as LEGAL-BERT-SC in Chalkidis et al. (2020); a model trained from scratch in the legal corpora mentioned below using a newly created vocabulary by a sentence-piece tokenizer trained on the very same corpora.
60
61 \*\* As many of you expressed interest in the LEGAL-BERT-FP models (those relying on the original BERT-BASE checkpoint), they have been released in Archive.org (https://archive.org/details/legal_bert_fp), as these models are secondary and possibly only interesting for those who aim to dig deeper in the open questions of Chalkidis et al. (2020).
62
63 ## Load Pretrained Model
64
65 ```python
66 from transformers import AutoTokenizer, AutoModel
67
68 tokenizer = AutoTokenizer.from_pretrained("nlpaueb/legal-bert-base-uncased")
69 model = AutoModel.from_pretrained("nlpaueb/legal-bert-base-uncased")
70 ```
71
72 ## Use LEBAL-BERT variants as Language Models
73
74 | Corpus | Model | Masked token | Predictions |
75 | --------------------------------- | ---------------------------------- | ------------ | ------------ |
76 | | **BERT-BASE-UNCASED** |
77 | (Contracts) | This [MASK] Agreement is between General Motors and John Murray . | employment | ('new', '0.09'), ('current', '0.04'), ('proposed', '0.03'), ('marketing', '0.03'), ('joint', '0.02')
78 | (ECHR) | The applicant submitted that her husband was subjected to treatment amounting to [MASK] whilst in the custody of Adana Security Directorate | torture | ('torture', '0.32'), ('rape', '0.22'), ('abuse', '0.14'), ('death', '0.04'), ('violence', '0.03')
79 | (EURLEX) | Establishing a system for the identification and registration of [MASK] animals and regarding the labelling of beef and beef products . | bovine | ('farm', '0.25'), ('livestock', '0.08'), ('draft', '0.06'), ('domestic', '0.05'), ('wild', '0.05')
80 | | **CONTRACTS-BERT-BASE** |
81 | (Contracts) | This [MASK] Agreement is between General Motors and John Murray . | employment | ('letter', '0.38'), ('dealer', '0.04'), ('employment', '0.03'), ('award', '0.03'), ('contribution', '0.02')
82 | (ECHR) | The applicant submitted that her husband was subjected to treatment amounting to [MASK] whilst in the custody of Adana Security Directorate | torture | ('death', '0.39'), ('imprisonment', '0.07'), ('contempt', '0.05'), ('being', '0.03'), ('crime', '0.02')
83 | (EURLEX) | Establishing a system for the identification and registration of [MASK] animals and regarding the labelling of beef and beef products . | bovine | (('domestic', '0.18'), ('laboratory', '0.07'), ('household', '0.06'), ('personal', '0.06'), ('the', '0.04')
84 | | **EURLEX-BERT-BASE** |
85 | (Contracts) | This [MASK] Agreement is between General Motors and John Murray . | employment | ('supply', '0.11'), ('cooperation', '0.08'), ('service', '0.07'), ('licence', '0.07'), ('distribution', '0.05')
86 | (ECHR) | The applicant submitted that her husband was subjected to treatment amounting to [MASK] whilst in the custody of Adana Security Directorate | torture | ('torture', '0.66'), ('death', '0.07'), ('imprisonment', '0.07'), ('murder', '0.04'), ('rape', '0.02')
87 | (EURLEX) | Establishing a system for the identification and registration of [MASK] animals and regarding the labelling of beef and beef products . | bovine | ('live', '0.43'), ('pet', '0.28'), ('certain', '0.05'), ('fur', '0.03'), ('the', '0.02')
88 | | **ECHR-BERT-BASE** |
89 | (Contracts) | This [MASK] Agreement is between General Motors and John Murray . | employment | ('second', '0.24'), ('latter', '0.10'), ('draft', '0.05'), ('bilateral', '0.05'), ('arbitration', '0.04')
90 | (ECHR) | The applicant submitted that her husband was subjected to treatment amounting to [MASK] whilst in the custody of Adana Security Directorate | torture | ('torture', '0.99'), ('death', '0.01'), ('inhuman', '0.00'), ('beating', '0.00'), ('rape', '0.00')
91 | (EURLEX) | Establishing a system for the identification and registration of [MASK] animals and regarding the labelling of beef and beef products . | bovine | ('pet', '0.17'), ('all', '0.12'), ('slaughtered', '0.10'), ('domestic', '0.07'), ('individual', '0.05')
92 | | **LEGAL-BERT-BASE** |
93 | (Contracts) | This [MASK] Agreement is between General Motors and John Murray . | employment | ('settlement', '0.26'), ('letter', '0.23'), ('dealer', '0.04'), ('master', '0.02'), ('supplemental', '0.02')
94 | (ECHR) | The applicant submitted that her husband was subjected to treatment amounting to [MASK] whilst in the custody of Adana Security Directorate | torture | ('torture', '1.00'), ('detention', '0.00'), ('arrest', '0.00'), ('rape', '0.00'), ('death', '0.00')
95 | (EURLEX) | Establishing a system for the identification and registration of [MASK] animals and regarding the labelling of beef and beef products . | bovine | ('live', '0.67'), ('beef', '0.17'), ('farm', '0.03'), ('pet', '0.02'), ('dairy', '0.01')
96 | | **LEGAL-BERT-SMALL** |
97 | (Contracts) | This [MASK] Agreement is between General Motors and John Murray . | employment | ('license', '0.09'), ('transition', '0.08'), ('settlement', '0.04'), ('consent', '0.03'), ('letter', '0.03')
98 | (ECHR) | The applicant submitted that her husband was subjected to treatment amounting to [MASK] whilst in the custody of Adana Security Directorate | torture | ('torture', '0.59'), ('pain', '0.05'), ('ptsd', '0.05'), ('death', '0.02'), ('tuberculosis', '0.02')
99 | (EURLEX) | Establishing a system for the identification and registration of [MASK] animals and regarding the labelling of beef and beef products . | bovine | ('all', '0.08'), ('live', '0.07'), ('certain', '0.07'), ('the', '0.07'), ('farm', '0.05')
100
101
102
103 ## Evaluation on downstream tasks
104
105 Consider the experiments in the article "LEGAL-BERT: The Muppets straight out of Law School". Chalkidis et al., 2020, (https://aclanthology.org/2020.findings-emnlp.261)
106
107 ## Author - Publication
108
109 ```
110 @inproceedings{chalkidis-etal-2020-legal,
111 title = "{LEGAL}-{BERT}: The Muppets straight out of Law School",
112 author = "Chalkidis, Ilias and
113 Fergadiotis, Manos and
114 Malakasiotis, Prodromos and
115 Aletras, Nikolaos and
116 Androutsopoulos, Ion",
117 booktitle = "Findings of the Association for Computational Linguistics: EMNLP 2020",
118 month = nov,
119 year = "2020",
120 address = "Online",
121 publisher = "Association for Computational Linguistics",
122 doi = "10.18653/v1/2020.findings-emnlp.261",
123 pages = "2898--2904"
124 }
125 ```
126
127 Ilias Chalkidis on behalf of [AUEB's Natural Language Processing Group](http://nlp.cs.aueb.gr)
128
129 | Github: [@ilias.chalkidis](https://github.com/seolhokim) | Twitter: [@KiddoThe2B](https://twitter.com/KiddoThe2B) |
130