📄 中文摘要
PyTorch 是一个广泛使用的深度学习框架，具有强大的张量操作能力。教程涵盖了张量的创建与检查、数据类型、设备与类型转换、索引、切片与布尔掩码等基础知识。进一步探讨了张量的重塑、视图与挤压，以及逐元素操作与归约。广播机制作为隐形的乘法器，简化了张量运算。矩阵乘法作为核心操作，深入分析了批量矩阵操作和爱因斯坦求和（einsum）方法。最后，线性代数的基本概念也被纳入讨论，为读者提供了全面的矩阵操作技能。通过本教程，读者能够从初学者成长为高级用户，掌握 PyTorch 中的矩阵操作技巧。
📄 English Summary
PyTorch Tensors & Matrix Operations: A Deep-Dive Tutorial
PyTorch is a widely used deep learning framework known for its powerful tensor manipulation capabilities. This tutorial covers the fundamentals of tensor creation and inspection, data types, devices, and casting, as well as indexing, slicing, and boolean masking. It further explores reshaping, views, and squeezing of tensors, along with element-wise operations and reductions. The broadcasting mechanism, acting as a silent multiplier, simplifies tensor computations. Matrix multiplication, a core operation, is analyzed in depth, including batched matrix operations and the Einstein summation (einsum) method. Additionally, basic concepts of linear algebra are discussed, providing readers with comprehensive skills in matrix operations. This tutorial enables readers to progress from beginners to advanced users, mastering matrix manipulation techniques in PyTorch.