Ricursive Intelligence 发布 Ripple-65B Preview，训练成本 3200 万美元
> we developed and trained Ripple-65B Preview for only $32M, which is one-tenth the cost of the previous open-weight SOTA on agentic coding¹
Ricursive Intelligence 发布 65B 稠密大模型 Ripple-65B Preview，从零预训练，称其在美国开源权重模型的智能体编码任务上达到 SOTA，超过 Muse Spark Open-816B、Beam-501B、Inkling-975B 和 Nemotron 3 Ultra-550B 等更大模型。swyx 引用其说明称，3200 万美元是从零开始的完整研发成本，涵盖预训练阶梯开发、数据配比、评测、调试、合成数据生成、后训练方案开发以及最终的预训练与后训练运行。
查看原帖今天我们发布 Ripple-65B Preview，这是我们从头预训练的 65B 稠密大语言模型。 Ripple 在美国开放权重模型的智能体编码任务上树立了新的最先进水平，显著超越了包括 Muse Spark Open-816B、Beam-501B、Inkling-975B 和 Nemotron 3 Ultra-550B 在内的更大规模模型。 了解更多：https://t.co/Pe04ZKhkE9
查看原文
Today we release Ripple-65B Preview, our 65B dense LLM we pretrained from scratch. Ripple sets a new state-of-the-art for U.S. open-weight models in agentic coding tasks, surpassing significantly larger models including Muse Spark Open-816B, Beam-501B, Inkling-975B, and Nemotron 3 Ultra-550B. Learn more: https://t.co/Pe04ZKhkE9
正文
> we developed and trained Ripple-65B Preview for only $32M, which is one-tenth the cost of the previous open-weight SOTA on agentic coding¹
this is like the avengers of open model SOTA
¹ "Note that $32M is our full R&D cost starting from scratch, including pretraining ladder development, data mixing, evals, debugging, synthetic data generation, posttraining recipe development, and the final hero pretraining and posttraining run."