1
Show HN: 3.125-Bit LLM quantization bypassing tensor cores
突破张量核心限制，实现3.125位LLM量化，大幅压缩模型却不损失精度。
Article URL: https://blog.djellalmohamedaniss.workers.dev/posts/data-free-3bit-quantization/ Comments URL: https://news.ycombinator.com/item?id=482205…