N

Nitibench Ccl Human Finetuned Bge M3

Developed by VISAI-AI
A fine-tuned version of BAAI/bge-m3 on Thai legal query data, supporting dense retrieval, lexical matching, and multi-vector interaction
Downloads 51
Release Time : 2/15/2025

Model Overview

This is a sentence embedding model optimized for Thai legal texts, particularly suitable for legal clause retrieval and similarity calculation tasks. The model was fine-tuned on the WangchanX-Legal-ThaiCCL-RAG dataset, improving performance in the legal domain.

Model Features

Multimodal Retrieval Capability
Supports three retrieval methods simultaneously: dense vector retrieval, lexical matching, and multi-vector interaction
Legal Domain Optimization
Specially fine-tuned for Thai legal texts, excelling in legal clause retrieval tasks
Automated Fine-tuning Process
Employs a fully automated data preparation and model fine-tuning pipeline to ensure model quality

Model Capabilities

Generate Text Embeddings
Calculate Sentence Similarity
Legal Clause Retrieval
Lexical Weight Analysis
Multi-vector Interaction Matching

Use Cases

Legal Information Retrieval
Legal Clause Matching
Automatically matches relevant legal clauses based on user queries
Achieves high accuracy with HR@10 0.938 on the NitiBench-CCL dataset
Tax Consultation Support
Assists tax consultation systems in providing precise regulatory references
HR@10 reaches 0.8 on the NitiBench-Tax dataset
Intelligent Customer Service
Legal Q&A System
Provides automated legal question answering for financial institutions
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
ÂĐ 2025AIbase