Logistic Regression script (without keras)

2b_LogisticRegression.py

@@ -0,0 +1,69 @@
+import tensorflow as tf
+import numpy as np
+import matplotlib.pyplot as plt
+
+
+print("Tensorflow version: {}".format(tf.__version__))
+print("Found GPU: {}".format(tf.test.gpu_device_name()))
+print("Executing eagerly: {}".format(tf.executing_eagerly()))
+
+assert tf.__version__ >= '2.0.0'
+
+
+
+LEARNING_RATE = 0.01
+NUM_EPOCHS = 25
+BATCH_SIZE = 100
+
+(train_X, train_Y), (test_X, test_Y) = tf.keras.datasets.mnist.load_data()
+train_X, test_X = train_X / 255.0, test_X / 255.0
+train_X = train_X.reshape(60000, -1)
+
+
+class Model(tf.keras.Model):
+
+    def __init__(self):
+        super(Model, self).__init__()
+        self.W = tf.Variable(initial_value=tf.random_normal_initializer, shape=(784, 10), dtype='float32')
+        self.b = tf.Variable(initial_value=tf.random_normal_initializer, shape=(10,), dtype='float32')
+
+    def call(self, inputs):
+        inputs = tf.keras.Input(shape=(784,), name="digits")
+        return tf.nn.softmax(tf.matmul(inputs, self.W) + self.b)
+
+
+def loss(model, inputs, targets):
+    prediction = model(inputs)
+    return tf.losses.softmax_cross_entropy(targets, logits=prediction)
+
+
+optimizer = tf.optimizers.Adam()
+
+
+def train(model, inputs, targets, learning_rate):
+    with tf.GradientTape() as tape:
+        loss_value = loss(model, inputs, targets)
+    dW, db = tape.gradient(loss_value, [model.W, model.b])
+    optimizer.apply_gradients(zip([dW, db], [model.W, model.b]))
+
+
+def accuracy(model, inputs, targets):
+    logits = model(inputs)
+    prediction = tf.argmax(logits, 1).numpy()
+    equality = tf.equal(prediction, tf.argmax(targets, 1))
+    return tf.reduce_mean(tf.cast(equality, tf.float32))
+
+
+model = Model()
+print(train_X.shape)
+tf.nn.softmax(tf.matmul(tf.cast(train_X, tf.float32), model.W) + model.b)
+
+for epoch in range(NUM_EPOCHS):
+    train(model, train_X, train_Y, learning_rate=0.1)
+    current_loss = loss(model, train_X, train_Y)
+
+    if epoch % 20 == 0:
+        print("Loss at step {:03d}: {:.3f}".format(epoch, current_loss))
+
+print("Final Loss: {}".format(accuracy(model, test_X, test_Y)))
+