Meta announced the release of four powerful new AI models and a series of responsible AI research results at Meta FAIR, aiming to promote open innovation and responsible development in the AI field. The models released this time cover multiple fields such as natural language processing, music generation and audio watermark detection, demonstrating Meta's latest progress in artificial intelligence technology and its commitment to benefiting the wider community. Meta emphasizes that its goal is to make state-of-the-art AI technology available to everyone, not just a few large tech companies.
Highlights of the new AI model are as follows:
Meta Chameleon: This is a series of 7B and 34B language models that support mixed-mode input and plain text output, bringing new possibilities to the field of natural language processing.
Meta multi-label prediction: This is an innovative pre-trained language model that uses multi-label prediction technology to predict multiple future words at the same time instead of just one, thus significantly improving the model's capabilities, training efficiency and prediction speed. .
Meta JASCO: This is a generative text-to-music model that can accept various conditional inputs, provide higher controllability, and bring new innovative tools to the field of music creation.
Meta AudioSeal: This is an audio watermark model designed for detecting AI-generated speech content and supports commercial licensing.
In addition to these models, Meta has released other Responsible Artificial Intelligence (RAI) results, including research, data, and code designed to measure and improve the representation of geographic and cultural preferences and diversity in AI systems.
Meta emphasizes that state-of-the-art AI should be accessible to everyone, not just a few big tech companies. The company is eager to see how the community will utilize these technologies.
The AI models and research results released by Meta not only demonstrate its strong strength in the field of artificial intelligence, but also reflect its commitment to responsible AI development. The openness of these models will promote further development and innovation of the AI community and promote artificial intelligence technology to better serve society.