aiqbench takeAI Products

This release advances efficient bidirectional encoding by adapting LFM2 decoders to maintain high speed on CPUs with an extended 8K context window. The architecture's ability to scale costs slowly for long inputs addresses a critical bottleneck for continuous inference tasks like intent routing and safety filtering without requiring GPUs. These models offer a practical path forward for deploying robust encoders in resource-constrained environments while preserving multilingual capabilities.

Source coveringMarkTechPost

What this isA model's reading of the article and related coverage, written for aiqbench. Not the publisher's words.

Liquid AI Releases LFM2.5-Encoder-230M and LFM2.5-Encoder-350M: Bidirectional Encoders That Stay Fast at 8K Context on CPU

MarkTechPostJuly 29, 20265 min read

Liquid AI released two open-weight bidirectional encoders, LFM2.5-Encoder-230M and LFM2.5-Encoder-350M. Both carry an 8,192-token context and are built on the LFM2 hybrid backbone. The 350M ranks fourth of 14 models on a 17-task GLUE, SuperGLUE, and multilingual suite, behind only larger models. The 230M clears one 8K-token forward pass on CPU in about 28 seconds. The post Liquid AI Releases LFM2.5-Encoder-230M and LFM2.5-Encoder-350M: Bidirectional Encoders That Stay Fast at 8K Context on CPU appeared first on MarkTechPost.

Liquid AI has released two open-weight bidirectional encoders, LFM2.5-Encoder-230M and LFM2.5-Encoder-350M. Both are masked language models built on the LFM2 hybrid backbone. Both carry an 8,192-token context. Encoders sit underneath classifiers, intent routers, safety filters, and PII detectors. Those jobs run continuously, usually without a GPU, and increasingly on longer inputs. BERT established the class. ModernBERT pushed its accuracy, speed, and context. Liquid AI’s argument is that the LFM2 architecture continues that line, because its cost grows more slowly as inputs get longer. The encoders are not trained from…

Excerpt shown under the publisher's RSS terms · 89 words · full text and figures are on the publisher's site

Read at MarkTechPost ↗Ask about this article