pytorch实现函数拟合

导入必要模块随机生成数据

import random
import numpy as np
import matplotlib.pyplot as plt
import torch
import torch.nn as nn
import torch.nn.functional as F

x_train_list = []
y_train_list = []
for i in range(1, 50):
    x = i*random.choice([0.7,0.8,0.9])
    y = i*random.choice([0.4,0.5,0.8,0.9])
    x_train_list.append(["%.2f" % x])
    y_train_list.append(["%.2f" % y])

x_train = np.array(x_train_list, dtype=np.float32) #将数据列表转为np.array
y_train = np.array(y_train_list, dtype=np.float32)

绘图以及将np.array转为tensor

plt.scatter(x_train,y_train)
plt.show()
x_train = torch.from_numpy(x_train)
y_train = torch.from_numpy(y_train)

模型构建以及 训练数据

 

class LinearRegression(nn.Module):
  def __init__(self):
    super(LinearRegression,self).__init__()
    self.linear = nn.Linear(1,10)