LocalAI has announced the release of several new AI models, enhancing its offerings in natural language processing and music generation. The new models include 'Dumpling-Qwen2.5-32B', 'deepseek-r1-qwen-2.5-32b-ablated', and 'l3.3-nevoria-r1-70b', each designed for specific applications such as dialogue and reasoning. Additionally, the 'confucius-o1-14b' model has been introduced as a lightweight reasoning tool. Another notable release is the 'FuseO1-DeepSeekR1-QwQ-32B-Preview', which integrates multiple open-source large language models to improve reasoning capabilities. Furthermore, a new open-source music generation model named 'YuE' has been launched, capable of creating full songs from lyrics, and is compatible with Hugging Face and LLAMA for easy fine-tuning. This model aims to rival existing platforms like Suno AI.
YuE: opensource model for full-song generation "We tackle the task of generating whole-song music audio from given lyrics, dubbed lyrics2song" boss, I'm scared https://t.co/wg6zz4n82J https://t.co/y2dIvkXrb3
WHAAAAT - meet YuE (乐) open-source full-song music generation model that rivals Suno AI Hugging Face & LLAMA-compatible for easy fine-tuning. Rebels around the world doin WORK https://t.co/yyq4S8aTuV
RIP Suno and Udio. 😬 China just dropped another open-source model: YuE, a full-song music generation model. From lyrics to complete songs, it can generate tracks up to minutes long. It’s Hugging Face and LLAMA-compatible, for easy fine-tuning. More details 👇 https://t.co/QrUY0IthHw