2D Max Pooling

题面

实现 2D 最大池化（max pooling）下采样：给定输入张量与 kernel_size/stride/padding，按窗口取最大值生成输出。

Implementation Requirements

• External libraries are not permitted
• The solve function signature must remain unchanged
• The final result must be stored in tensor output

Examples

Example 1: 单通道 3×3 输入，kernel=2, stride=1 → 输出 2×2
Example 2: 5×5 输入，kernel=3, stride=1, padding=1 → 输出与页面示例一致

Constraints

• 1 ≤ N ≤ 100 (batch size)
• 1 ≤ C ≤ 512 (channels)
• 1 ≤ H, W ≤ 1024
• 1 ≤ kernel_size ≤ 16
• 1 ≤ stride ≤ 16
• 0 ≤ padding ≤ 16
• 输入/输出均为 float32；Performance: N = 4, kernel=3, stride=2