kobigbird-base23-84859751 Open-source Korean Q&A Model - Accurately Answer Korean Questions!

Home

Kobigbird Base23 84859751

Developed by sudong97

Korean Q&A model fine-tuned on the custom_squad_v2 dataset based on monologg/kobigbird-bert-base

Question Answering System

Transformers

#Korean Q&A System #Long Text Processing #BigBird Architecture

Downloads 13

Release Time : 1/9/2023

Model Overview

This model is a fine-tuned version of a Korean pre-trained model based on the BigBird architecture, specializing in Q&A tasks and optimized on a custom SQuAD-formatted dataset

Model Features

Long Text Processing Capability

Based on the BigBird architecture, excels at handling Q&A tasks with long sequence texts

Korean Language Optimization

Specifically pre-trained and fine-tuned for Korean text

Efficient Training

Utilizes optimization techniques like gradient accumulation and cosine annealing learning rate scheduling

Model Capabilities

Korean Text Comprehension

Q&A Generation

Long Document Processing

Use Cases

Customer Service

Korean FAQ System

Used to build an automatic Q&A system for Korean frequently asked questions

Education

Korean Learning Assistance

Helps Korean learners understand texts and answer questions

Training Loss	Epoch	Step	Validation Loss
No log	0.99	42	1.6141
No log	1.99	84	1.4628

Featured Recommended AI Models

Empowering the Future, Your AI Solution Knowledge Base

English 简体中文繁體中文にほんご

Kobigbird Base23 84859751

Model Overview

Model Features

Model Capabilities

Use Cases

🚀 kobigbird-base23-84859751

🚀 Quick Start

📚 Documentation

Model description

Intended uses & limitations

Training and evaluation data

🔧 Technical Details

Training procedure

Training hyperparameters

Training results

Framework versions