#低比特训练

1 stories · 1 in 7 days · 1 in 30 days
Stories, products and papers tagged "低比特训练", newest first.
Nov 14
ModelsBuildersMulti-source05:27
Chinese contentThinking Machines 公布 Inkling 2 Small 全程原生 4-bit 训练细节

Thinking Machines 把 Inkling 2 Small 从预训练到 RL rollout 全程用原生 4-bit 前向跑完,速度更快但量化误差会伤学习,这篇博客讲了他们怎么解决。做训练或想省算力成本的工程师值得看看细节。

Event
Official accountThinking Machines Lab@thinkymachinesSource