crazty的博客

pytorchPyTorch由4个主要包组成：torch：类似于Numpy的通用数组库，可将张量类型转换torch.cuda.TensorFloat，并在GPU上进行计算。torch.autograd：用于构建计算图形并自动获取梯度的包。torch.nn：具有共享层和损失函数的神经网络库。torch.optim：具有通用优化算法（如SGD，Adam等）的优化包接口torch.func...

官网nn.AdaptiveAvgPool2dnn.AdaptiveAvgPool2d就是自适应平均池化，指定输出（H，W）m = nn.AdaptiveAvgPool2d((5,1))input = torch.randn(1, 64, 8, 9)output = m(input)print(output.shape)torch.Size([1, 64, 5, 1])m = nn.AdaptiveAvgPool2d((None,1))input = torch.randn(1, 64,

pytorch的模型保存和读取torch.save: Saves a serialized object to disk. This function uses Python’s pickle utility for serialization. Models, tensors, and dictionaries of all kinds of objects can be saved using this function.torch.load: Uses pickle’s unpickling