OpenViSTAI · Agents · Blockchain · Crypto
Submit a project
Menu
AI Models

Kosentencebert Etri

Kosentencebert Etri is a Python implementation of Korean Sentence-BERT using Siamese ETRI KoBERT to generate sentence embeddings for Korean text.

Explore the practical guide ↓

Independent OpenViST project profile, based on linked sources.

Kosentencebert Etri — OpenViST project profile in AI Models.
OpenViST project illustration. See the linked sources for product details.
Understand the project

Practical overview

When it fits

Use it when building a Korean text-processing workflow that needs sentence vectors for semantic matching or clustering, and you can manage the model assets and modified Python dependencies.

Who it is for

  • Developers building Korean semantic search or sentence-similarity features.
  • NLP practitioners who want to cluster Korean sentences using embedding vectors.
  • Teams training sentence-embedding models with NLI or STS data.

Documented uses

  • Encode Korean corpus sentences and rank them against a query by cosine similarity.
  • Group sentences using their embedding similarity, with the README showing KMeans as an example.
  • Train models using NLI data, STS data, or NLI followed by STS fine-tuning.
Before you choose

Requirements & limits

  • The repository modifies Transformers, Sentence Transformers and tokenizers code and recommends an isolated virtual environment.
  • ETRI KoBERT and tokenizer files are required but are not included in this repository.
  • The README specifies an ETRI KoBERT compatibility range of Transformers 2.4.1–2.8.0 and says Sentence-BERT requires version 3.1.0 or above.
  • Training requires KorNLUDatasets in the project directory.
  • The documented pooling strategy is mean pooling; the README reports test-set correlation metrics for several distance measures.

Check current requirements in the official documentation. Source-based guidance does not mean OpenViST has run or independently tested this project.

Source-based practical guide

Prepare the repository environment

Set up the documented Python environment and place the required model and tokenizer files for use.

Before you start

  • Python and pip
  • ETRI KoBERT model and tokenizer files, which are not provided in this repository

Steps

  1. Clone the repository and create a Python virtual environment.

  2. Activate the environment and install the listed requirements.

  3. Move the Transformers, tokenizers and Sentence Transformers directories into the environment's site-packages location as described in the README.

  4. Ensure the ETRI KoBERT model and tokenizer are in the expected KoSentenceBERT directory.

Expected outcome

The documented project environment and required ETRI model assets are in place for subsequent model use.

Guide limitations

  • The README's example site-packages path refers to Python 3.7.
  • This repository does not provide ETRI KoBERT or its tokenizer.
  • The instructions describe setup, not a verified execution or deployment.
Documentation 1 ↗
Evidence and transparency

Sources & review

Official repository ↗
Source record details
Source availability
Active
Sources checked
18 October 2026
Original website title
BM-K/KoSentenceBERT-ETRI · Official homepage title · retained in its source language
How we review sourcesSuggest a correction

Related options

KoSentenceBERT_SKTBERT — The README identifies this as a separate version using SKT KoBERT.