I

Indobart V2

Developed by indobenchmark
IndoBART-v2 is an Indonesian language model based on the BART architecture, specifically designed for natural language generation tasks.
Downloads 983
Release Time : 3/2/2022

Model Overview

IndoBART-v2 is a pretrained Indonesian language model trained using BART's training objectives, suitable for various natural language processing tasks, especially text generation.

Model Features

BART-based Architecture
Utilizes BART's training objectives, combining the strengths of bidirectional encoders and autoregressive decoders.
Large-scale Pretraining Data
Pretrained using the Indo4B-Plus dataset (26 GB of text).
Indonesian Language Optimization
Specifically optimized for Indonesian, suitable for Indonesian natural language processing tasks.

Model Capabilities

Text Generation
Text Summarization
Machine Translation
Question Answering

Use Cases

Natural Language Generation
Indonesian Text Summarization
Generates concise summaries of Indonesian texts.
Indonesian Machine Translation
Translates other languages into Indonesian.
Question Answering
Indonesian Question Answering
Answers questions based on Indonesian texts.
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
Š 2025AIbase