創作 pytorch不同的層設定不同的學習率

import torch

from torch import nn, optim

from torch.autograd import Variable

import numpy as np

import matplotlib.pyplot as plt

x_train = np.array([[3.3], [4.4], [5.5], [6.71], [6.93], [4.168],

                   [9.779], [6.182], [7.59], [2.167], [7.042],

                   [10.791], [5.313], [7.997], [3.1]], dtype=np.float32)

y_train = np.array([[1.7], [2.76], [2.09], [3.19], [1.694], [1.573],

                   [3.366], [2.596], [2.53], [1.221], [2.827],

                   [3.465], [1.65], [2.904], [1.3]], dtype=np.float32)

x_train = torch.from_numpy(x_train)

y_train = torch.from_numpy(y_train)

# Linear Regression Model

class LinearRegression(nn.Module):

   def __init__(self):

       super(LinearRegression, self).__init__()

       self.linear1 = nn.Linear(1, 5)  # input and output is 1 dimension

       self.linear2 = nn.Linear(5, 1)

   def forward(self, x):

       out = self.linear1(x)

       out = self.linear2(out)

       return out

model = LinearRegression()

print(model.linear1)

# 微調：自定義每一層的學習率

# 定義loss和優化函數

criterion = nn.MSELoss()

optimizer = optim.SGD(

   [{"params": model.linear1.parameters(), "lr": 0.01},

    {"params": model.linear2.parameters()}],