TensorFlow2.0代码案例

原创于 2020-06-08 17:39:07 发布 · 2k 阅读

0 ·

本内容遵循CC 4.0 BY-SA版权协议

智能钻完井同时被 2 个专栏收录

253 篇文章

订阅专栏

Python在石油工程中应用

81 篇文章

订阅专栏

本文通过使用TensorFlow和NumPy库，演示了如何基于随机生成的数据集训练一个简单的线性回归模型。通过调整权重和偏置，模型能够拟合输入数据与目标输出之间的线性关系。

import tensorflow as tf
import numpy as np
input_xs = np.random.rand(1000)
input_ys = 3 * input_xs + 0.217
weight = tf.Variable(1., dtype=tf.float32, name="weight")
bias = tf.Variable(1., dtype=tf.float32, name="bias")

def model(xs):
    logits = tf.multiply(xs, weight) + bias
    return logits
opt = tf.optimizers.Adam(learning_rate=1e-4)
for xs, ys in zip(input_xs, input_ys):
    xs = np.reshape(xs, [1])
    ys = np.reshape(ys, [1])
    with tf.GradientTape() as tape:
        _loss = tf.reduce_mean(tf.pow((model(xs) - ys), 2)) / (2 * 1000)
    grads = tape.gradient(_loss, [weight, bias])
    opt.apply_gradients(zip(grads, [weight, bias]))
    print('Training loss is :', _loss.numpy())
print(weight)
print(bias)

作者：ChenBD