Model Overview Card for Qwen Max 2025-01-25

Basic Information

• Model Name: Qwen Max 2025-01-25
• Developer/Creator: Alibaba
• Release Date: January 25, 2025
• Version: Qwen2.5
• Model Type: Text

Description

Overview:

Qwen Max is a large-scale Mixture-of-Experts (MoE) language model excelling in chat, coding, and problem-solving.

Key Features:

• Mixture-of-Experts (MoE) architecture
• Pre-trained on over 20 trillion tokens
• Supervised Fine-Tuning (SFT) and Reinforcement Learning from Human Feedback (RLHF)
• 32K context window
• Multi-language support (English and Chinese)

Intended Use:

Chat-based interactions, code generation, complex question answering.

Language Support:

English and Chinese

Technical Details

Architecture:

Mixture-of-Experts (MoE). The model's parameters are divided into "expert" blocks, with each input token routed to a subset of these experts. It utilizes a Transformer-based architecture.

Training Data:

• Data Source and Size: Pre-trained on over 20 trillion tokens, including multi-lingual textual data and domain-specific corpora (scientific literature, code).
• Diversity and Bias: Includes significant coverage of English and Chinese textual data, along with coverage of other major world languages.

Performance Metrics

‍

Usage

Code Samples:

The model is available on the AI/ML API platform as "Qwen Max 2025-01-25" .

API Documentation:

Detailed API Documentation is available here.

Ethical Guidelines

The Qwen team emphasizes ethical considerations in AI development by promoting transparency regarding the model's capabilities and limitations. The organization encourages responsible usage to prevent misuse or harmful applications of generated content.

Licensing

Qwen Max is available under specific licensing terms provided by Alibaba Cloud. Users are advised to review the licensing information to understand the permissions and restrictions associated with the model's use.

‍

Get Qwen Max 2025-01-25 API here.