Merge pull request #1 from vbvg2008/master

updated FE 1.0 working examples

image_classification/lenet_mnist.py

@@ -0,0 +1,46 @@
+from fastestimator.pipeline.static.preprocess import Minmax
+from fastestimator.estimator.estimator import Estimator
+from fastestimator.pipeline.pipeline import Pipeline
+from fastestimator.architecture.lenet import LeNet
+from fastestimator.estimator.trace import Accuracy
+import tensorflow as tf
+import numpy as np
+
+class Network:
+    def __init__(self):
+        self.model = LeNet()
+        self.optimizer = tf.optimizers.Adam()
+        self.loss = tf.losses.SparseCategoricalCrossentropy()
+
+    def train_op(self, batch):
+        with tf.GradientTape() as tape:
+            predictions = self.model(batch["x"])
+            loss = self.loss(batch["y"], predictions)
+        gradients = tape.gradient(loss, self.model.trainable_variables)
+        self.optimizer.apply_gradients(zip(gradients, self.model.trainable_variables))
+        return predictions, loss
+
+    def eval_op(self, batch):
+        predictions = self.model(batch["x"], training=False)
+        loss = self.loss(batch["y"], predictions)
+        return predictions, loss
+
+def get_estimator(epochs=2, batch_size=32, optimizer="adam"):
+
+    (x_train, y_train), (x_eval, y_eval) = tf.keras.datasets.mnist.load_data()
+    x_train = np.expand_dims(x_train, -1)
+    x_eval = np.expand_dims(x_eval, -1)
+
+    pipeline = Pipeline(batch_size=batch_size,
+                        feature_name=["x", "y"],
+                        train_data={"x": x_train, "y": y_train},
+                        validation_data={"x": x_eval, "y": y_eval},
+                        transform_train= [[Minmax()], []])
+
+    traces = [Accuracy(feature_true="y")]
+
+    estimator = Estimator(network= Network(),
+                          pipeline=pipeline,
+                          epochs= epochs,
+                          traces= traces)
+    return estimator

image_generation/dcgan_mnist.py

@@ -0,0 +1,102 @@
+from fastestimator.estimator.estimator import Estimator
+from fastestimator.pipeline.pipeline import Pipeline
+from tensorflow.keras import layers
+import tensorflow as tf
+import numpy as np
+
+class Network:
+    def __init__(self):
+        self.discriminator = self.make_discriminator_model()
+        self.generator = self.make_generator_model()
+        self.cross_entropy = tf.keras.losses.BinaryCrossentropy(from_logits=True)
+        self.generator_optimizer = tf.keras.optimizers.Adam(1e-4)
+        self.discriminator_optimizer = tf.keras.optimizers.Adam(1e-4)
+
+    def make_generator_model(self):
+        model = tf.keras.Sequential()
+        model.add(layers.Dense(7*7*256, use_bias=False, input_shape=(100,)))
+        model.add(layers.BatchNormalization())
+        model.add(layers.LeakyReLU())
+        model.add(layers.Reshape((7, 7, 256)))
+        assert model.output_shape == (None, 7, 7, 256) # Note: None is the batch size
+        model.add(layers.Conv2DTranspose(128, (5, 5), strides=(1, 1), padding='same', use_bias=False))
+        assert model.output_shape == (None, 7, 7, 128)
+        model.add(layers.BatchNormalization())
+        model.add(layers.LeakyReLU())
+        model.add(layers.Conv2DTranspose(64, (5, 5), strides=(2, 2), padding='same', use_bias=False))
+        assert model.output_shape == (None, 14, 14, 64)
+        model.add(layers.BatchNormalization())
+        model.add(layers.LeakyReLU())
+        model.add(layers.Conv2DTranspose(1, (5, 5), strides=(2, 2), padding='same', use_bias=False, activation='tanh'))
+        assert model.output_shape == (None, 28, 28, 1)
+        return model
+
+    def make_discriminator_model(self):
+        model = tf.keras.Sequential()
+        model.add(layers.Conv2D(64, (5, 5), strides=(2, 2), padding='same', input_shape=[28, 28, 1]))
+        model.add(layers.LeakyReLU())
+        model.add(layers.Dropout(0.3))
+        model.add(layers.Conv2D(128, (5, 5), strides=(2, 2), padding='same'))
+        model.add(layers.LeakyReLU())
+        model.add(layers.Dropout(0.3))
+        model.add(layers.Flatten())
+        model.add(layers.Dense(1))
+        return model
+
+    def discriminator_loss(self, real_output, fake_output):
+        real_loss = self.cross_entropy(tf.ones_like(real_output), real_output)
+        fake_loss = self.cross_entropy(tf.zeros_like(fake_output), fake_output)
+        total_loss = real_loss + fake_loss
+        return total_loss
+
+    def generator_loss(self, fake_output):
+        return self.cross_entropy(tf.ones_like(fake_output), fake_output)
+
+    def train_op(self, batch):
+        noise = tf.random.normal([32, 100])
+        with tf.GradientTape() as gen_tape, tf.GradientTape() as disc_tape:
+            generated_images = self.generator(noise, training=True)
+
+            real_output = self.discriminator(batch["x"], training=True)
+            fake_output = self.discriminator(generated_images, training=True)
+
+            gen_loss = self.generator_loss(fake_output)
+            disc_loss = self.discriminator_loss(real_output, fake_output)
+
+        gradients_of_generator = gen_tape.gradient(gen_loss, self.generator.trainable_variables)
+        gradients_of_discriminator = disc_tape.gradient(disc_loss, self.discriminator.trainable_variables)
+
+        self.generator_optimizer.apply_gradients(zip(gradients_of_generator, self.generator.trainable_variables))
+        self.discriminator_optimizer.apply_gradients(zip(gradients_of_discriminator, self.discriminator.trainable_variables))
+        return generated_images, (gen_loss, disc_loss)
+
+    def eval_op(self, batch):
+        noise = tf.random.normal([32, 100])
+        generated_images = self.generator(noise, training=False)
+        real_output = self.discriminator(batch["x"], training=False)
+        fake_output = self.discriminator(generated_images, training=False)
+        gen_loss = self.generator_loss(fake_output)
+        disc_loss = self.discriminator_loss(real_output, fake_output)
+        return generated_images, (gen_loss, disc_loss)
+
+class Myrescale:
+    def transform(self, data, decoded_data=None):
+        data = tf.cast(data, tf.float32)
+        data = (data - 127.5) / 127.5
+        return data
+
+def get_estimator():
+    (x_train, _), (x_eval, _) = tf.keras.datasets.mnist.load_data()
+    x_train = np.expand_dims(x_train, -1)
+    x_eval = np.expand_dims(x_eval, -1)
+
+    pipeline = Pipeline(batch_size=32,
+                        feature_name=["x"],
+                        train_data={"x": x_train},
+                        validation_data={"x": x_eval},
+                        transform_train= [[Myrescale()], []])
+
+    estimator = Estimator(network= Network(),
+                          pipeline=pipeline,
+                          epochs= 2)
+    return estimator