AIbase
EN
Home
Acge Text Embedding
A

Acge Text Embedding

Developed by aspire
The ACGE Text Embedding Model is designed for generating text embeddings and supports various natural language processing tasks.
Text Embedding
Safetensors
#Chinese Semantic Similarity#Medical QA Re-ranking#Multi-task Text Embedding
Downloads 27.12k
Release Time : 3/9/2024

Model Overview

This model primarily generates high-quality text embeddings, suitable for tasks such as semantic similarity calculation, text classification, clustering, retrieval, and re-ranking.

Model Features

Multi-task Support
Supports various natural language processing tasks, including semantic similarity calculation, text classification, clustering, retrieval, and re-ranking.
High Performance
Delivers outstanding performance on multiple benchmark datasets, especially in Chinese text processing tasks.

Model Capabilities

Text Embedding Generation
Semantic Similarity Calculation
Text Classification
Text Clustering
Text Retrieval
Text Re-ranking

Use Cases

Semantic Similarity Calculation
Sentence Similarity Calculation
Calculates the semantic similarity between two sentences, applicable in scenarios like QA systems and recommendation systems.
On the AFQMC dataset, the Pearson correlation for cosine similarity is 54.03.
Text Classification
Amazon Review Classification
Performs sentiment classification on Amazon product reviews.
Accuracy is 48.54%, and F1 score is 46.60%.
Text Clustering
Sentence Clustering
Clusters semantically similar sentences together.
On the CLSClusteringP2P dataset, the V-measure is 47.08%.
Text Retrieval
Medical QA Retrieval
Retrieves relevant answers from medical QA datasets.
On the Cmedqa retrieval dataset, MAP@10 is 40.00%.
Text Re-ranking
Medical QA Re-ranking
Re-ranks retrieval results to improve relevance.
On the CMedQAv1 dataset, MAP is 88.66%.
Featured Recommended AI Models
AIbase
Empowering the Future, Your AI Solution Knowledge Base
English简体中文繁體中文にほんご
© 2025AIbase