Model Overview Card for Voyage Large 2 Instruct

Basic Information

Model Name: Voyage Large 2 Instruct

Developer/Creator: Voyage AI

Release Date: May 2024

Version: 2.0

Model Type: Text Embedding Model

Description

Overview:

Voyage Large 2 Instruct is an instruction-tuned, general-purpose text embedding model optimized for tasks such as clustering, classification, and retrieval. It is designed to perform exceptionally well on the Massive Text Embedding Benchmark (MTEB), ranking first in several key areas.

Key Features:

• Instruction tuning for enhanced task performance
• Optimized for clustering, classification, and retrieval
• High performance on MTEB benchmarks
• 16K context window for improved understanding

Intended Use:

Voyage Large 2 Instruct is intended for use in scenarios requiring high-quality text embeddings, such as:

• Text retrieval and search
• Text classification and clustering
• Reranking tasks

Language Support:

The model primarily supports English but can be adapted for other languages with appropriate tuning.

Technical Details

Architecture:

Voyage Large 2 Instruct is based on a transformer architecture, which is commonly used in modern natural language processing tasks due to its efficiency and scalability.

Training Data:

The model was trained on a diverse dataset that includes a wide range of text sources. This diversity helps ensure robust performance across different tasks and reduces potential biases.

Data Source and Size:

The training data is extensive, leveraging millions of text samples to cover various domains and topics. This large dataset size contributes to the model's comprehensive understanding of language.

Diversity and Bias:

Efforts have been made to include diverse data sources to minimize bias, though the model's effectiveness can vary depending on the specific application and data it encounters.

Performance Metrics

Comparison to Other Models:

Accuracy:

Voyage Large 2 Instruct demonstrates high accuracy across various tasks, outperforming many competitors in classification and retrieval tasks.

Speed:

The model is optimized for a balance between speed and accuracy, making it suitable for real-time applications.

Robustness:

It handles diverse inputs well and generalizes effectively across different topics and languages, thanks to its extensive training data.

Usage

Sign in to get immediate access to Voyage Large 2 Instruct and schedule a call to discuss your use case with us.

Code Samples:

Ethical Guidelines:

Voyage Large 2 Instruct is developed with ethical considerations in mind, aiming to minimize biases and ensure fair use across applications. Developers are encouraged to assess the model's performance in their specific contexts to address any ethical concerns.

License Type:

The model is available under a commercial license, allowing for both commercial and non-commercial use, subject to the terms specified by Voyage AI.