# RAG Optimization

This is a very compact and fast Japanese reranking model, suitable for improving the accuracy of RAG systems.

The Nomic code embedding model is a top-tier code retrieval tool that supports multiple programming languages and excels in code retrieval tasks.

Chonky is a Transformer model capable of intelligently splitting text into meaningful semantic chunks, suitable for RAG systems.

This is a fine-tuned sentence-transformers model specifically optimized for financial document retrieval tasks while maintaining general performance.

Chonky is a Transformer model that intelligently splits text into meaningful semantic chunks for RAG systems.

Chonky is a Transformer model that intelligently segments text into meaningful semantic chunks, suitable for RAG systems.

A 7-billion-parameter multimodal embedding model specialized in visual document retrieval tasks, achieving outstanding performance on the Vidore-v2 benchmark

Multilingual embedding model for German and English, supporting a context length of 8192 tokens

A Chinese text chunking tool based on BertForTokenClassification, particularly suitable for processing unstructured and messy text

A question classifier fine-tuned based on IBM Granite embedding model, used to distinguish between general questions and directed questions

A BERT-based text chunker that predicts chunk start markers via a classifier head, employing sliding window technology to handle documents of any length, suitable for both structured and unstructured text.

A model specifically designed for Arabic reranking tasks, capable of accurately processing the relationship between queries and passages, directly evaluating the similarity between questions and documents, and outputting relevance scores.

This is the Investopedia embedding model developed by the FinLang team for financial applications. It is fine-tuned based on BAAI/bge-base-en-v1.5 and maps sentences and paragraphs into a 768-dimensional dense vector space, suitable for semantic search and other tasks in the financial domain.

BCEmbedding is a bilingual and cross-lingual embedding model library developed by NetEase Youdao, including EmbeddingModel (semantic vector generation) and RerankerModel (result reranking). As a core component of Youdao's Retrieval-Augmented Generation (RAG) system, it has been successfully applied to open-source projects like QAnything and products such as Youdao Speed Reading and Youdao Translation.

A Reranker model fine-tuned on Korean data based on BAAI/bge-reranker-large, designed to enhance Korean Retrieval-Augmented Generation (RAG) performance

Inkbot is a conversational AI model specifically designed for parsing and responding to structured prompts, supporting dynamic dialogues both with and without context.