Korean-Sentence-Embedding

Korean sentence embedding repository. You can download the pre-trained models and inference right away, also it provides environments where individuals can train models.

Quick tour

Note
All the pretrained models are uploaded in Huggingface Model Hub. Check https://huggingface.co/BM-K

import torch
from transformers import AutoModel, AutoTokenizer

def cal_score(a, b):
    if len(a.shape) == 1: a = a.unsqueeze(0)
    if len(b.shape) == 1: b = b.unsqueeze(0)

    a_norm = a / a.norm(dim=1)[:, None]
    b_norm = b / b.norm(dim=1)[:, None]
    return torch.mm(a_norm, b_norm.transpose(0, 1)) * 100

model = AutoModel.from_pretrained('BM-K/KoSimCSE-roberta-multitask')  # or 'BM-K/KoSimCSE-bert-multitask'
tokenizer = AutoTokenizer.from_pretrained('BM-K/KoSimCSE-roberta-multitask')  # or 'BM-K/KoSimCSE-bert-multitask'

sentences = ['치타가 들판을 가로 질러 먹이를 쫓는다.',
             '치타 한 마리가 먹이 뒤에서 달리고 있다.',
             '원숭이 한 마리가 드럼을 연주한다.']

inputs = tokenizer(sentences, padding=True, truncation=True, return_tensors="pt")
embeddings, _ = model(**inputs, return_dict=False)

score01 = cal_score(embeddings[0][0], embeddings[1][0])  # 84.09
# '치타가 들판을 가로 질러 먹이를 쫓는다.' @ '치타 한 마리가 먹이 뒤에서 달리고 있다.'
score02 = cal_score(embeddings[0][0], embeddings[2][0])  # 23.21
# '치타가 들판을 가로 질러 먹이를 쫓는다.' @ '원숭이 한 마리가 드럼을 연주한다.'

Update history

** Updates on Mar.08.2023 **

Update Unsupervised Models

** Updates on Feb.24.2023 **

Upload KoSimCSE clustering example

** Updates on Nov.15.2022 **

Upload KoDiffCSE-unsupervised training code

** Updates on Oct.27.2022 **

Upload KoDiffCSE-unsupervised performance

** Updates on Oct.21.2022 **

Upload KoSimCSE-unsupervised performance

** Updates on Jun.01.2022 **

Release KoSimCSE-multitask models

** Updates on May.23.2022 **

Upload KoSentenceT5 training code
Upload KoSentenceT5 performance

** Updates on Mar.01.2022 **

Release KoSimCSE

** Updates on Feb.11.2022 **

Upload KoSimCSE training code
Upload KoSimCSE performance

** Updates on Jan.26.2022 **

Upload KoSBERT training code
Upload KoSBERT performance

Baseline Models

Baseline models used for korean sentence embedding - KLUE-PLMs

Model	Embedding size	Hidden size	# Layers	# Heads
KLUE-BERT-base	768	768	12	12
KLUE-RoBERTa-base	768	768	12	12

Warning
Large pre-trained models need a lot of GPU memory to train

Available Models

Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks [SBERT]-[EMNLP 2019]
SimCSE: Simple Contrastive Learning of Sentence Embeddings [SimCSE]-[EMNLP 2021]
Sentence-T5: Scalable Sentence Encoders from Pre-trained Text-to-Text Models [Sentence-T5]-[ACL findings 2022]
DiffCSE: Difference-based Contrastive Learning for Sentence Embeddings [DiffCSE]-[NAACL 2022]

Datasets

kakaobrain KorNLU Datasets (Supervised setting)
wiki-corpus (Unsupervised setting)

Setups

KoSentenceBERT

🤗 Model Training
Dataset (Supervised)
- Training: snli_1.0_train.ko.tsv, sts-train.tsv (multi-task)
  - Performance can be further improved by adding multinli data to training.
- Validation: sts-dev.tsv
- Test: sts-test.tsv

KoSimCSE

🤗 Model Training
Dataset (Supervised)
- Training: snli_1.0_train.ko.tsv + multinli.train.ko.tsv (Supervised setting)
- Validation: sts-dev.tsv
- Test: sts-test.tsv
Dataset (Unsupervised)
- Training: wiki_corpus.txt
- Validation: sts-dev.tsv
- Test: sts-test.tsv

KoSentenceT5

🤗 Model Training
Dataset (Supervised)
- Training: snli_1.0_train.ko.tsv + multinli.train.ko.tsv
- Validation: sts-dev.tsv
- Test: sts-test.tsv

KoDiffCSE

🤗 Model Training
Dataset (Unsupervised)
- Training: wiki_corpus.txt
- Validation: sts-dev.tsv
- Test: sts-test.tsv

Performance-supervised

Model	Average	Cosine Pearson	Cosine Spearman	Euclidean Pearson	Euclidean Spearman	Manhattan Pearson	Manhattan Spearman	Dot Pearson	Dot Spearman
KoSBERT^†_SKT	77.40	78.81	78.47	77.68	77.78	77.71	77.83	75.75	75.22
KoSBERT	80.39	82.13	82.25	80.67	80.75	80.69	80.78	77.96	77.90
KoSRoBERTa	81.64	81.20	82.20	81.79	82.34	81.59	82.20	80.62	81.25

KoSentenceBART	77.14	79.71	78.74	78.42	78.02	78.40	78.00	74.24	72.15
KoSentenceT5	77.83	80.87	79.74	80.24	79.36	80.19	79.27	72.81	70.17

KoSimCSE-BERT^†_SKT	81.32	82.12	82.56	81.84	81.63	81.99	81.74	79.55	79.19
KoSimCSE-BERT	83.37	83.22	83.58	83.24	83.60	83.15	83.54	83.13	83.49
KoSimCSE-RoBERTa	83.65	83.60	83.77	83.54	83.76	83.55	83.77	83.55	83.64

KoSimCSE-BERT-multitask	85.71	85.29	86.02	85.63	86.01	85.57	85.97	85.26	85.93
KoSimCSE-RoBERTa-multitask	85.77	85.08	86.12	85.84	86.12	85.83	86.12	85.03	85.99

Performance-unsupervised

Model	Average	Cosine Pearson	Cosine Spearman	Euclidean Pearson	Euclidean Spearman	Manhattan Pearson	Manhattan Spearman	Dot Pearson	Dot Spearman
KoSRoBERTa-base^†	N/A	N/A	48.96	N/A	N/A	N/A	N/A	N/A	N/A
KoSRoBERTa-large^†	N/A	N/A	51.35	N/A	N/A	N/A	N/A	N/A	N/A

KoSimCSE-BERT	74.08	74.92	73.98	74.15	74.22	74.07	74.07	74.15	73.14
KoSimCSE-RoBERTa	75.27	75.93	75.00	75.28	75.01	75.17	74.83	75.95	75.01

KoDiffCSE-RoBERTa	77.17	77.73	76.96	77.21	76.89	77.11	76.81	77.74	76.97

Korean-SRoBERTa^†

Downstream tasks

KoSBERT: Semantic Search, Clustering
KoSimCSE: Semantic Search, Clustering

License

This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.

References

@misc{park2021klue,
    title={KLUE: Korean Language Understanding Evaluation},
    author={Sungjoon Park and Jihyung Moon and Sungdong Kim and Won Ik Cho and Jiyoon Han and Jangwon Park and Chisung Song and Junseong Kim and Yongsook Song and Taehwan Oh and Joohong Lee and Juhyun Oh and Sungwon Lyu and Younghoon Jeong and Inkwon Lee and Sangwoo Seo and Dongjun Lee and Hyunwoo Kim and Myeonghwa Lee and Seongbo Jang and Seungwon Do and Sunkyoung Kim and Kyungtae Lim and Jongwon Lee and Kyumin Park and Jamin Shin and Seonghyun Kim and Lucy Park and Alice Oh and Jung-Woo Ha and Kyunghyun Cho},
    year={2021},
    eprint={2105.09680},
    archivePrefix={arXiv},
    primaryClass={cs.CL}
}

@inproceedings{gao2021simcse,
   title={{SimCSE}: Simple Contrastive Learning of Sentence Embeddings},
   author={Gao, Tianyu and Yao, Xingcheng and Chen, Danqi},
   booktitle={Empirical Methods in Natural Language Processing (EMNLP)},
   year={2021}
}

@article{ham2020kornli,
  title={KorNLI and KorSTS: New Benchmark Datasets for Korean Natural Language Understanding},
  author={Ham, Jiyeon and Choe, Yo Joong and Park, Kyubyong and Choi, Ilji and Soh, Hyungjoon},
  journal={arXiv preprint arXiv:2004.03289},
  year={2020}
}

@inproceedings{reimers-2019-sentence-bert,
    title = "Sentence-BERT: Sentence Embeddings using Siamese BERT-Networks",
    author = "Reimers, Nils and Gurevych, Iryna",
    booktitle = "Proceedings of the 2019 Conference on Empirical Methods in Natural Language Processing",
    month = "11",
    year = "2019",
    publisher = "Association for Computational Linguistics",
    url = "http://arxiv.org/abs/1908.10084",
}

@inproceedings{chuang2022diffcse,
   title={{DiffCSE}: Difference-based Contrastive Learning for Sentence Embeddings},
   author={Chuang, Yung-Sung and Dangovski, Rumen and Luo, Hongyin and Zhang, Yang and Chang, Shiyu and Soljacic, Marin and Li, Shang-Wen and Yih, Wen-tau and Kim, Yoon and Glass, James},
   booktitle={Annual Conference of the North American Chapter of the Association for Computational Linguistics (NAACL)},
   year={2022}
}