https://thinkless-github-io.pages.dev/tags/%E6%A8%A1%E5%9E%8B%E6%9E%B6%E6%9E%84/Recent content in 模型架构 on 杨の草原Hugozh-CNTue, 06 May 2025 21:25:21 +0800https://thinkless-github-io.pages.dev/posts/%E7%BB%8F%E5%85%B8%E6%A8%A1%E5%9E%8B%E4%B8%8E%E6%9E%B6%E6%9E%84/Tue, 06 May 2025 21:25:21 +0800https://thinkless-github-io.pages.dev/posts/%E7%BB%8F%E5%85%B8%E6%A8%A1%E5%9E%8B%E4%B8%8E%E6%9E%B6%E6%9E%84/大模型经典架构笔记，记录多模态模型连接方式、视觉编码器选择、文本解码器设计和 DeepSeek 技术细节。https://thinkless-github-io.pages.dev/posts/transformer%E6%A8%A1%E5%9D%97/Tue, 29 Apr 2025 09:32:21 +0800https://thinkless-github-io.pages.dev/posts/transformer%E6%A8%A1%E5%9D%97/Transformer 架构笔记，记录 Encoder-Decoder 结构、多头自注意力、前馈网络、残差连接和层归一化等基础模块。