Alibaba releases 100 open-source AI models and new text-to-video generator
Alibaba Cloud launched over 100 open-source AI models in the Qwen 2.5 family, including a text-to-video generator and the Qwen2-VL model for advanced video comprehension, enhancing global AI infrastructure.
Read original article
Alibaba Cloud has announced the release of over 100 new open-source artificial intelligence models as part of the Qwen 2.5 family, unveiled at the Apsara Conference. This follows the success of the Tongyi Qianwen (Qwen) foundation model, which has been downloaded over 40 million times. The new models vary in size from 500 million to 72 billion parameters, allowing for a range of applications from simple tasks to complex language understanding. Alibaba has also introduced enhancements in coding and mathematics capabilities within these models. Additionally, a new text-to-video generator has been launched, capable of producing high-quality videos from prompts in both Chinese and English, as well as transforming static images into videos. This model utilizes advanced diffusion transformer architecture to improve video quality. The company is heavily investing in AI technology and infrastructure to support global customers, as stated by Eddie Wu, CEO of Alibaba Cloud Intelligence. The recent developments also include the Qwen2-VL model, which offers advanced vision comprehension and can understand videos up to 20 minutes long, designed for integration into various devices.
- Alibaba Cloud released over 100 open-source AI models in the Qwen 2.5 family.
- The models range from 500 million to 72 billion parameters, catering to various tasks.
- A new text-to-video generator can create videos from prompts and enhance static images.
- Alibaba is focusing on building AI infrastructure to support global business needs.
- The Qwen2-VL model offers advanced video comprehension and is suitable for mobile and automotive integration.
Related
China Is Closing the A.I. Gap with the United States
Chinese tech companies are advancing in AI, showcasing innovations like Kuaishou's video generator. Despite U.S. trade restrictions, they leverage open-source technologies, supported by the government, to enhance development and competition.
Apple Intelligence Foundation Language Models
Apple has developed language models to enhance its Apple Intelligence features, including a compact on-device model and a larger server-based model, emphasizing Responsible AI and improving user interactions in iOS and macOS.
CogVideoX: A Cutting-Edge Video Generation Model
ZhipuAI launched CogVideoX, an advanced video generation model featuring a 3D Variational Autoencoder for efficient data compression and an end-to-end understanding model, enhancing video generation and instruction responsiveness.
Show HN: Infinity – Realistic AI characters that can speak
Infinity AI has developed a groundbreaking video model that generates expressive characters from audio input, trained for 11 GPU years at a cost of $500,000, addressing limitations of existing tools.
Qwen2.5: A Party of Foundation Models
Qwen has released Qwen2.5, a major update featuring specialized models for coding and mathematics, pretrained on 18 trillion tokens, supporting long text generation and multilingual capabilities across 29 languages.
0 commentsRelated
China Is Closing the A.I. Gap with the United States
Chinese tech companies are advancing in AI, showcasing innovations like Kuaishou's video generator. Despite U.S. trade restrictions, they leverage open-source technologies, supported by the government, to enhance development and competition.
Apple Intelligence Foundation Language Models
Apple has developed language models to enhance its Apple Intelligence features, including a compact on-device model and a larger server-based model, emphasizing Responsible AI and improving user interactions in iOS and macOS.
CogVideoX: A Cutting-Edge Video Generation Model
ZhipuAI launched CogVideoX, an advanced video generation model featuring a 3D Variational Autoencoder for efficient data compression and an end-to-end understanding model, enhancing video generation and instruction responsiveness.
Show HN: Infinity – Realistic AI characters that can speak
Infinity AI has developed a groundbreaking video model that generates expressive characters from audio input, trained for 11 GPU years at a cost of $500,000, addressing limitations of existing tools.
Qwen2.5: A Party of Foundation Models
Qwen has released Qwen2.5, a major update featuring specialized models for coding and mathematics, pretrained on 18 trillion tokens, supporting long text generation and multilingual capabilities across 29 languages.