Encoder Decoder和decoder Only架构训练和推理浅析
LLM时代的transformer参数量、计算量、激活值的分析
A survey of Efficient Transformer on Inference
[cuda-learning-notes] 硬件抽象和执行模型
[cuda-learning-notes] 内存模型
[cuda-learning-notes] 流和同步
[cuda-learning-notes] 工具使用和profile
[cuda-learning-notes] GPU架构发展、兼容性和编译
[Effective Modern Cpp Notes] Ch08 微调
[Effective Modern Cpp Notes] Ch07 并发API
Effective Modern Cpp Reading Notes
[Effective Modern Cpp Notes] Ch06 Lambda表达式
[Effective Modern Cpp Notes] Ch05 右值引用、移动语句和完美转发
[Effective Modern Cpp Notes] Ch04 智能指针
[Effective Modern Cpp Notes] Ch03 转向现代C++
[Effective Modern Cpp Notes] Ch02 Auto
[Effective Modern Cpp Notes] Ch01 类型推导
[Effective Cpp Notes] Ch08 定制new和delete
[Effective Cpp Notes] Ch07 模板与泛型编程
[Effective Cpp Notes] Ch06 继承与面向对象设计
[Effective Cpp Notes] Ch05 实现
[Effective Cpp Notes] Ch04 设计与声明
[Effective Cpp Notes] Ch03 资源管理
[Effective Cpp Notes] Ch02 构造、析构、赋值运算
[Effective Cpp Notes] Ch01 让自己习惯C++