Model Overview Card for Magnum V4 72B

Basic Information

• Model Name: Magnum V4 72B
• Developer/Creator: Anthracite Organization
• Release Date: October September 202420, 2024
• Version: 0.1
• Model Type: Large Language Model (LLM)

Description

Overview:

Magnum V4 72B is a large language model fine-tuned on top of Qwen2.5, specifically designed to replicate the prose quality of the Claude 3 models, particularly Sonnet and Opus. It excels in generating coherent and contextually rich text, making it suitable for various applications requiring high-quality language generation.

Key Features:

• 72 billion parameters for advanced language understanding and generation.
• Supports a context length of up to 32,768 tokens.
• Instruction-tuned for improved performance on specific tasks.
• Incorporates advanced tokenization techniques with a vocabulary size of 152,064.
• Multi-language support covering languages such as English, French, German, Spanish, Italian, Portuguese, Russian, Chinese, and Japanese.

Intended Use:

This model is designed for applications such as content creation, automated customer support, interactive storytelling, and any scenario requiring high-quality conversational AI.

Language Support:

Magnum V4 supports multiple languages, enhancing its applicability in global contexts.

Technical Details

Architecture:

Magnum V4 is based on the Transformer architecture, specifically utilizing the Qwen2ForCausalLM framework. Key architectural details include:

• Layers: 40
• Hidden Dimension: 14,336
• Number of Heads: 32
• Activation Function: GELU
• Data Type: bfloat16 for efficient computation.

Training Data:

The model was trained on a diverse dataset that includes conversational logs and instructional data to enhance its dialogue capabilities.

• Data Source and Size: The training dataset consists of various sources including chat logs and instructional texts aimed at improving the model's ability to engage in conversation effectively.
• Knowledge Cutoff: The model's knowledge is current as of September 2024.
• Diversity and Bias: The training data was curated to minimize bias while maximizing diversity in topics and dialogue styles. This approach enhances the model's ability to handle varied inputs effectively.

Performance Metrics:

Magnum V4 has shown competitive performance across various benchmarks:

• Achieves high accuracy in generating contextually appropriate responses.
• Strong performance in maintaining coherence over long dialogues.

Usage

Code Samples:

The model is available on the AI/ML API platform as "Magnum v4 72B" .

API Documentation:

Detailed API Documentation is available here.

Ethical Guidelines

Anthracite Organization emphasizes ethical considerations in AI development by promoting transparency regarding the model's capabilities and limitations. They encourage users to adhere to responsible usage guidelines to prevent misuse or harmful applications.

Licensing

Magnum V4 is licensed under a custom license allowing non-commercial usage while restricting commercial applications without permission.

‍

Get Magnum v4 72B API here.