Pytorch学习笔记

最新推荐文章于 2025-05-24 14:25:39 发布

Dini_Dini

最新推荐文章于 2025-05-24 14:25:39 发布

阅读量116

点赞数

文章标签：深度学习

本文链接：https://blog.youkuaiyun.com/qq_44625074/article/details/113484641

版权

梯度下降入门

import torch
def gradDescent(X, y, eps = torch.tensor(0.01, requires_grad = True), numIt = 1000):
    m, n = X.shape
    weights = torch.zeros(n, 1, requires_grad = True)
    for k in range(numIt):
        grad = torch.mm(X.t(), (torch.mm(X, weights) - y))/2
        weights = weights - eps * grad
    return weights
X = torch.tensor([[1.,1],[3, 1]], requires_grad = True)
y = torch.tensor([2.,4], requires_grad = True).reshape(2,1)
weights = gradDescent(X, y, numIt = 10000)
print(weights)