Japanese SimCSE Technical Report

Hayato Tsukagoshi, Ryohei Sasano, Koichi Takeda

10/30/2023

cs.CL

No official code found. Best implementation: hppRC/simple-simcse-ja (69 stars, PyTorch, last commit 2023-10-31).

Abstract

We report the development of Japanese SimCSE, Japanese sentence embedding models fine-tuned with SimCSE. Since there is a lack of sentence embedding models for Japanese that can be used as a baseline in sentence embedding research, we conducted extensive experiments on Japanese sentence embeddings involving 24 pre-trained Japanese or multilingual language models, five supervised datasets, and four unsupervised datasets. In this report, we provide the detailed training setup for Japanese SimCSE and their evaluation results.

View on arXiv View PDF

Code Implementations(1)

hppRC/simple-simcse-jaCommunity100%

PyTorch

Exploring Japanese SimCSE

694Shell, PythonNov 11, 20222 years ago

japanesepytorchsentence-embeddingssentence-transformerssimcse+1 more

Similar Papers

Translation Aligned Sentence Embeddings for Turkish Language

91% match2023

Scaling Sentence Embeddings with Large Language Models

88% match1 repo1092023

Sentence Embedding Models for Ancient Greek Using Multilingual Knowledge Distillation

87% match1 repo162023

Word Embeddings Are Steers for Language Models

86% match1 repo1472023

JCSE: Contrastive Learning of Japanese Sentence Embeddings and Its Applications

86% match1 repo82023

Training Effective Neural Sentence Encoders from Automatically Mined Paraphrases

86% match1 repo222022

Cite this paper

@article{tsukagoshi2023japanese,
  title  = {Japanese SimCSE Technical Report},
  author = {Hayato Tsukagoshi and Ryohei Sasano and Koichi Takeda},
  year   = {2023},
  eprint = {2310.19349},
  archivePrefix = {arXiv},
  url    = {http://arxiv.org/abs/2310.19349v1}
}

Add a badge to your README

Shows whether official code is linked for this paper.

[![Code of Paper](https://codeofpaper.com/badge/2310.19349.svg)](https://codeofpaper.com/paper/2310.19349)