Deepseek training 2T and plans 8T model
英文标题：Deepseek training 2T and plans 8T model
Quote: DeepSeek is training a 2T-parameter model and plans to eventually build an 8T-parameter model. https://x.com/wallstengine/status/2101982843656388644 Current DeepSeek models: Flash parameter count of 552 billion Pro: 1.6T (trillion) total parameters with 49B (billion) activated weights per to
打开原文相关 2 · 机会 0
中文
Quote: DeepSeek is training a 2T-parameter model and plans to eventually build an 8T-parameter model. https://x.com/wallstengine/status/2101982843656388644 Current DeepSeek models: Flash parameter count of 552 billion Pro: 1.6T (trillion) total parameters with 49B (billion) activated weights per to
中文全文翻译正在补齐，可先查看下方英文原文。
English / 原文
Quote: DeepSeek is training a 2T-parameter model and plans to eventually build an 8T-parameter model. https://x.com/wallstengine/status/2101982843656388644 Current DeepSeek models: Flash parameter count of 552 billion Pro: 1.6T (trillion) total parameters with 49B (billion) activated weights per to