Alibaba releases latest Qwen AI platform

- Advertisement -

ALIBABA released Qwen 2.5, just at about the same time Hangzhou DeepSeek Artificial Intelligence launched DeepSeek. This latest version of its artificial intelligence platform is a large language model (LLM) builds upon the foundation of Tongyi Qianwen, which was first introduced in April 2023.

According to Alibaba, Qwen 2.5 represents a significant leap forward, showcasing advancements in size, architecture, and capabilities. One notable aspect of the platform is its expanded model range, offering versions with varying parameter sizes, from 0.5 billion to a massive 72 billion parameters. This allows for greater flexibility in deployment, catering to both resource-constrained environments like mobile devices and demanding enterprise-level applications.  

Alibaba has invested heavily in training Qwen 2.5, utilizing a massive 18-trillion-token dataset. This extensive training data contributes to the model’s improved understanding and response accuracy across various tasks. Furthermore, it boasts an extended context window, capable of processing up to 128,000 tokens. This enables the model to handle more complex and nuanced prompts, generating more detailed and comprehensive responses.

- Advertisement -

Alibaba views Qwen 2.5 as a strategic asset in its pursuit of AI leadership. But compared to DeepSeek it isn’t fully open-source, but offers partial open-source access, allowing developers to fine-tune and adapt the model to their specific needs. It may open be to collaboration but not in the same depth as DeepSeek’s R1.

With Qwen 2.5, Alibaba is not just keeping pace with the rapid advancements in AI, but actively shaping the future of the technology. The company’s ongoing investment in research and development, coupled with its focus on accessibility and practical applications, positions Qwen 2.5 as a key player in the evolving landscape of large language models.

Author

- Advertisement -

Share post: