LiquidAI has released two new multilingual bidirectional encoders, LFM2.5-Encoder-230M and LFM2.5-Encoder-350M, built on the LFM2 architecture. These models are designed for on-device efficiency and fine-tuning for various downstream tasks across 15 languages. The smaller 230M model is optimized for low latency and memory, while the larger 350M model aims for maximum quality, both offering an 8k context window and competitive performance. AI
IMPACT These models offer efficient, on-device multilingual capabilities, potentially accelerating adoption in applications requiring low latency and broad language support.
RANK_REASON New model release from a known AI lab (LiquidAI) with detailed technical specifications and performance claims. [lever_c_demoted from frontier_release: ic=2 ai=1.0]
Read on Hugging Face Trending Models →
- Google Colab
- Hugging Face
- Kaggle
- LFM2
- LFM2.5-Encoder
- Lfm2BidirectionalModel
- LiquidAI/LFM2.5-Encoder-230M
- LiquidAI/LFM2.5-Encoder-350M
- ModernBERT
- transformers
AI-generated summary · Google Gemini · from 2 sources. How we write summaries →
