TII releases Falcon-Edge series of 1.58-bit BitNet language models in 1B and 3B sizes
Original title:Falcon-Edge: A series of powerful, universal, fine-tunable 1.58bit language models.
AI summary
TII's Falcon-Edge series is based on the BitNet architecture, pre-trained directly with ternary weights {-1, 0, 1}, and comes in 1B and 3B parameter sizes, each with base and instruct versions. The models were pre-trained on an internal data mix of about 1.5 Tera Tokens.
Key Takeaways
- TII's Falcon-Edge series is based on the BitNet architecture,pre-trained directly with ternary…
- The models were pre-trained on an internal data mix of about 1.
- 5 Tera Tokens.
Quick Facts
- Published
- Updated
- Source
- Hugging Face 社区博客(混合发现)
- Category
- Model Releases
- AI score
- 64
原文解释了 BitNet 三值权重的训练范式和微调路径,读者可以据此判断 1.58bit 模型在端侧部署的适用性。
Source & copyright
This article is based on curated titles and AI summaries from the AIHOT public API, with attribution and links back. Original copyright remains with the original publishers and authors; this site does not mirror full texts.
How to Cite
When citing this article, keep the title, date, and permalink, and credit Orange News as the source.
Orange News: TII releases Falcon-Edge series of 1.58-bit BitNet language models in 1B and 3B sizes. May 15, 2025. https://new.nat6.net/en/news/fm425yt03kg5jtbfgp99um7jm/